&EPA
          United States
          Environmental Protection
          Agency
         Office of Air Quality
         Planning and Standards
         Research Triangle Park NC 27711
March 1988
          Air
Hazardous Waste
TSDF-Background
Information for
Proposed RCRA
Air Emission
Standards
    Draft
    EIS
          Volume l-Chapters
          PRELIMINARY DRAFT

-------
                            NOTICE
This document has not been formally released by EPA and should not now be construed to represent Agency policy. It is being
circulated for comment on its technical accuracy and policy implications.
        Hazardous Waste TSDF-Background
        Information for Proposed  RCRA Air
                  Emission Standards

                   Volume l-Chapters

                PRELIMINARY DRAFT
                       Emission Standards Division
                   U. S. ENVIRONMENTAL PROTECTION AGENCY
                        Office of Air and Radiation
                    Office of Air Quality Planning and Standards
                   Research Triangle Park, North Carolina 27711

                           March 1988

-------
                                CONTENTS


Chapter



1.0
2.0
3.0









4.0


















Figures 	
Tables 	 	
Abbreviations and Conversion Factors 	
Introduction 	 	
Regulatory Authority and Standards Development 	
Industry Description and Air Emissions 	
3.1 The Hazardous Waste Industry 	
311 General Hazardous Waste Description 	
3.1.2 Generators 	
3.1.3 Transporters 	
3.1.4 Treatment, Storage, and Disposal Facilities...
3.1.5 TSDF Emission Sources 	
3.2 Estimates of Organic Emissions 	
3.2.1 Emission Estimation Data Requirements 	
3.2.2 Nationwide TSDF Emissions 	
3.3 References 	
Control Technologies 	
4.1 Suppression Controls 	
4.1.1 Enclosures 	
4.1.2 Covers 	
4.2 Add-on Emission Control Devices 	
4.2.1 Generic Control Devices 	
4.2.2 Carbon Adsorption 	
4.2.3 Combust i on 	
4.2.4 Condensation 	
4.3 Organic Removal Processes and Technologies 	
4.3.1 Steam Stripping 	
4 3.2 Air Stripping 	
4.3.3 Thin-Film Evaporation 	
4 3.4 Batch Distillation 	
4 3.5 Dewateri ng 	
4.4 Hazardous Waste Incineration 	
4.5 Coking of Petroleum Refinery Wastes 	
4.6 Process Modifications and Improved Work
Practices 	

                                                                         VI
                                                                        vi i

                                                                         ix
                                                                        3-1
                                                                        3-1
                                                                        3-1
                                                                        3-4
                                                                        3-5
                                                                        3-7
                                                                       3-14
                                                                       3-18
                                                                       3-19
                                                                       3-21
                                                                       3-30

                                                                        4-1
                                                                        4-3
                                                                        4-3
                                                                        4-9
                                                                       4-18
                                                                       4-18
                                                                       4-19
                                                                       4-23
                                                                       4-31
                                                                       4-35
                                                                       4-35
                                                                       4-41
                                                                       4-45
                                                                       4-49
                                                                       4-54
                                                                       4-57
                                                                       4-61

                                                                       4-63
'This  portion  of the document is currently under development.
                                   iii

-------
                           CONTENTS  (continued)
Chapter
Page
               4.6.1  Submerged Loading	   4-63
               4.6.2  Subsurface Injection	   4-64
               4.6.3  Daily Earth Covers	   4-65
               4.6.4  Mechanical Mixers	   4-65
               4.6.5  Housekeeping in Drum Storage Areas	   4-67
          4.7  Control of Equipment Leak Emissions from
               Waste Transfer	   4-68
          4.8  Cross-Media and Secondary Environmental
               Impacts	   4-69
          4.9  Summary of Controls Selected for Development
               of Control Strategies	   4-70
          4.10 References	   4-78

  5.0     Control Strategies	    5-1
          5.1  Control Strategy Concept	    5-1
               5.1.1  Combinations of Emission Sources,
                      Controls, and Cutoffs	    5-2
               5.1.2  Approaches to Selecting Control
                      Strategies for Evaluation	    5-5
          5.2  Example Control  Strategies	    5-7
          5.3  Impacts To Be Estimated for Control Strategies	    5-9
               5.3.1  List of Nationwide Impacts to Be
                      Estimated	    5-9
               5.3.2  Baseline for Nationwide Impacts
                      Estimates	   5-10
          5.4  Reference		   5-14

  6.0     National Organic Emissions and Health Risk Impacts	    6-1
          6.1  Organic Emission Impacts	    6-1
          6.2  Human Health Risks	    6-4
               6.2.1  Annual Cancer Incidence	    6-7
               6.2.2  Maximum Lifetime Risk	   6-10
               6.2.3  Noncancer Health Effects Assessment--
                      Acute and Chronic Exposures	   6-10
          6.3  Other Environmental  Impacts	      *

  7.0     National Control  Costs	    7-1
          7.1  Control  Costs Development	    7-1
               7.1.1  Methodology for Model  Units, Organic
                      Removal,  and Waste Incineration
                      Control  Costs	    .7-2
               7.1.2  Derivation of Unit Costs to Estimate
                      Nationwide Costs of Example Control
                      Strategies	    7-4
          7.2  Summary of Nationwide Control  Costs for Control
               Strategies	    7-6
          7.3  Cost  Effectiveness of Control  Strategies	    7-9
          7.4  References	   7-11
^This  portion  of the document is currently under development.
                                   IV

-------
                           CONTENTS (continued)





Appendix  (bound separately in Volume II)                              Page



  A       Evolution of Proposed Standards	    A-l



  B       Index to Environmental Impact Considerations	    B-l



  C       Emission Models and Emission Estimates	    C-l



  D       Source Assessment Model	    D-l



  E       Estimating Health Effects	    E-l



  F       Test Data	    F-l



  G       Emission Measurement and Continuous Monitoring	,	    G-l



  H       Costing of Add-On and Suppression Controls	    H-l



  I       Costing of Organic Removal  Processes and



          Hazardous Waste Incineration	    1-1



  J       Exposure Assessment for Maximum Risk and Noncancer



          Health Effects	    J-l

-------
                                  FIGURES
Number                                          -                       Page

  3-1      Simplified hazardous  waste  system  from generation  to
          disposal	     3-2
  3-2      Estimate  of physical  characteristics  of RCRA
          hazardous wastes	     3-6
  3-3      Two examples of onsite  hazardous waste land treatment
          operations	    3-10
  3-4      Two examples of active  landfill operations	    3-11
  3-5      Example onsite  hazardous  waste  storage facility	    3-12
  3-6      Source Assessment  Model  (SAM)  input files  used  in
          estimating nationwide treatment, storage,  and
          disposal  facilities  (TSDF)  uncontrolled air emissions	    3-25

  4-1      Typical air-supported structure 	     4-4
  4-2      Storage tank covers	    4-10
  4-3      Nonregenerative carbon  adsorption  (carbon
          canister)  unit  	    4-20
  4-4      Schematic diagram  of  thermal  incinerator system	    4-24
  4-5      Schematic diagram  of  catalytic  incinerator system  	    4-27
  4-6      Steam-assisted  elevated flare  system  	    4-30
  4-7      Schematic diagram  of  a  shell-and tube-heat surface
          condenser 	    4-33
  4-8      Schematic diagram  of  a  steam  stripping system	    4-36
  4-9      Schematic diagram  of  an air stripping  system	    4-42
  4-10     Schematic diagram  of  a  thin-film evaporator system 	    4-47
  4-11     Schematic diagram  of  batch distillation with
          fractionating column  	    4-51
  4-12     Dewatering  system  with  enclosed dewatering device  	    4-56
  4-13     Hazardous  waste incinerators  	    4-60
                                  VI

-------
                                  TABLES
Number
  3-1     Resource Conservation and Recovery Act (RCRA)
          Hazardous Waste Management Definitions	     3-8
  3-2     Nationwide Quantity of Hazardous Waste Managed
          by Specific Processes	     3-9
  3-3     Hazardous Waste Management Process Emission Sources	    3-15
  3-4     Summary of Selected Model Hazardous Waste Management
          Unit Uncontrolled Organic Emission Estimates for
          Model Wastes	    3-22
  3-5     Nationwide Uncontrolled TSDF Organic Emission
          Estimates	    3-29

  4-1     Potential Organic Air Emission Reduction Option
          for TSDF Sources	     4-2
  4-2     Emission Control Options Used for Selecting
          TSDF Control Strategies	    4-72
  4-3     Emission Control Efficiencies Used in
          Estimating Nationwide Impacts of Control Strategies	    4-74
  4-4     Generic Control Device Definitions	    4-77

  5-1     TSDF Emission Source Categories 	     5-3
  5-2     Potential Source/Control/Cutoff Combinations	     5-6
  5-3     Example TSDF Air Emission Control  Strategies	     5-8
  5-4     Nationwide Impacts to Be Estimated for Control
          Strategies	    5-11

  6-1     Summary of Nationwide Organic Emissions  and Health
          Risk Impacts for Uncontrolled, Baseline, and Two
          Example Control Strategies	     6-2
  6-2     Nationwide TSDF Emissions and Emission Reduction
          for the Uncontrolled, Baseline, and Two  Example
          Control  Strategies	     6-5
  6-3     Nationwide.Cancer Incidence from TSDF Emissions by
          Source Category	     6-8
  6-4     Maximum Lifetime Risks from TSDF Emissions	    6-11
                                   VI 1

-------
                            TABLES (continued)
Number

  7-1


  7-2


  7-3
Estimated Total Capital Investment and Total
Annual Cost Per Unit of Waste Throughput by
Source Category for Two Example Control Strategies
Estimated Nationwide Total Capital Investment and
Total Annual Cost for Two Example Control
Strategies	
Nationwide TSDF Cost Effectiveness of Two Example
Control  Strategies	°.	
 7-7


 7-8

7-10
                                 VI 1 1

-------
                  ABBREVIATIONS AND CONVERSION FACTORS
     The EPA policy is to express all measurements in Agency documents in
the International  System of Units (SI).  Listed below are abbreviations
and conversion factors for equivalents of these units.
Abbreviations

L - liters


kg - kilograms


Mg - megagrams
m   meters
cm - centimeters

kPa -  kilopascals
ha   hectares


rad - radians


kW - kilowatts
                Conversion Factor

                liter X 0.26   = gallons
                gallons X 3.79 = liters

                kilograms X 2.203 - pounds
                pounds X 0.454    - kilograms

                megagram XI       = metric tons
                megagram X 1.1     = short tons
                short tons X 0.907 = megagrams

                meters X 3.28       = feet
                centimeters X 0.396 = inches

                kilopascals X 0.01 = bars
                bars X 100         = kilopascals
                kilopascals X 0.0099 = atmospheres
                atmospheres X 101    = kilopascals
                kilopascals X 0.145 = pound per
                  square inch
                pound per square inch X 6.90 =
                  kilopascals

                hectares X 2.471 = acres
                acres X 0.40469  = hectares
                radians X 0.1592
                revolutions X 6.281

                kilowatts X 1.341
                horsepower X 0.7457
                    revolutions
                    radians

                    horsepower
                    ki1owatts
          Frequently used measurements in this document are:
                 0.21
                 5.7
                30
                76
               800
                 1.83
m3
m3
m3
m3
            210 L
          5,700 L
     ~   30,000 L
     ~   76,000 L
m^   ~  800,000 L
kg 02/kW/h
kW/28,.3 m3
kPa»m3/g«mol
          55 gal
       1,500 gal
       8,000 gal
      20,000 gal
  ~  210,000 gal
 3 Ib 02/hp/h
1.341 hp/103 ft3
0.0099 atm«m3/g»mol
                                   IX

-------
                3.0  INDUSTRY DESCRIPTION AND AIR EMISSIONS

     This chapter presents a brief overview of the hazardous waste industry
and a summary description of the techniques used in estimating nationwide
organic air emissions for hazardous waste treatment, storage, and disposal
facilities (TSDF) in the United States.  The hazardous waste industry and
TSDF emission sources are described in Section 3.1.  The estimation of TSDF
nationwide emissions is presented in Section 3.2.  Emission estimation
techniques include the development and use of (1) TSDF emission models,
which provide a mechanism for analyzing air emissions from TSDF management
processes and applicable emission control technologies,  and (2) a computer
program developed to process the data and information on the TSDF industry
and to perform emission calculations based on the available data.  Discus-
sions of air pollution controls and control strategies at TSDF follow in
Chapters 4.0 and 5.0, respectively.
3.1  THE HAZARDOUS WASTE INDUSTRY
     The hazardous waste industry in the United States is diverse and com-
plex.  The universe of hazardous waste generators represents a broad spec-
trum of industry types and sizes.  Wastes generated vary considerably in
both composition and form; and the waste management processes and practices
used in treating, storing, and disposing of hazardous wastes are also
widely varied.  Figure 3-1 presents a simplified waste system flow chart
for the hazardous waste industry.  Key elements of the industry are:  gene-
ration,  transportation.,  treatment, storage, and disposal.  The major ele-
ments of the hazardous waste industry are discussed in the following sec-
tions.
3.1.1  General Hazardous Waste Description
     General  waste descriptions include hazardous wastes in the following
forms:   contaminated wastewaters, spent solvents residuals, still bottoms,
                                    3-1

-------
      Generation
Transportation
Storage
Treatment
Disposal
CO
I

RCRA Wastes
Large & oman
Ljuaniiiieb






































INTERFIRM or
INTRAFIRM
Land & Water
Transport




































^



















Drum Tank
& Impound-
ment Storage























































—





Commercial
Incinerators*


Onsite
Incinerators*


Solvent &
Other
Recovery
Operations

Other
Operations'!"

Industrial
Furnaces
or Boilers*
























^



















Commercial
Disposal


Onsite
Disposal


Deep Well
Injection






    *Regulated by the Office of Solid Waste.
    tOnsite and commercial.
                              Figure 3-1.  Simplified hazardous waste system from generation to disposal.

-------
spent catalysts, electroplating wastes, metal-contaminated sludges,
degreasing solvents, leaded tank bottoms, American Petroleum Institute
(API) separator sludges, off-specification chemicals, and a variety of
other waste types.   In  reviewing waste data, more than 4,000 chemical con-
stituents have been  identified as being contained in the various waste
types examined.^
     Title 40 of the Code of Federal Regulations  (CFR),  Part 261.3  (40 CFR
261.3), defines hazardous waste as four categories:
     •    Characteristic wastes—wastes that exhibit any hazardous
          characteristic identified in 40 CFR 261 Subpart C, includ-
          ing:  ignitibi1ity, corrosivity, reactivity, or extraction
          procedure  (EP) toxicity
          Listed waste—wastes listed  in 40 CFR 261, Subpart D
     •    Mixture  rule  wastes—wastes  that are  (1) a mixture of solid
          waste and  a characteristic waste unless the mixture no
          longer exhibits any hazardous characteristic,  or (2) a mix-
          ture of  a  solid waste and one or more listed hazardous
          wastes
     •    Derived  from  rule wastes — any solid waste generated from  the
          treatment, storage, or disposal of a  hazardous waste,
          including  any sludge, spill  residue,  ash, emission control
          dust, or leachate  (but not including  precipitation runoff).
     Hazardous wastes are designated by Resource  Conservation and Recovery
Act  (RCRA) alphanumeric codes.  Codes  D001 through D017 are referred to as
"characteristic wastes."  D001 represents wastes  that are ignitible in
character; D002, those  that are corrosive; and  D003, those that are reac-
tive.  Extracts of wastes that contain toxic concentrations of specific
metals, pesticides,  or  herbicides are  assigned  one of the codes D004
through D017.
     "Listed wastes" encompass four groups of alphanumeric codes published
in 40 CFR 261, Subpart  D.  Hazardous wastes generated from nonspecific
industry sources such as degreasing operations  and electroplating are
listed as codes beginning with the letter "F,"  e.g., FOOl.  Hazardous
wastes from specific generating sources such as petroleum refining  are  '
assigned codes beginning with the letter "K," e.g., K048.  Waste codes
beginning with "P" or "U" represent waste commercial chemical products and
manufacturing chemical   intermediates (whether usable or off-specification)
                                    3-3

-------
     40 CFR 261,  "Identification and Listing of Hazardous Wastes," not only
 lists  hazardous wastes but also identifies specific wastes that are
 excluded  from regulation as hazardous.  These excluded wastes can be
 stored, treated,  or disposed of without a RCRA permit.
 3.1.2  Generators
     The  overwhelming majority of hazardous wastes are produced by large-
 quantity  generators, those firms that generate more than 1,000 kg of
 hazardous waste per month.  It has been estimated that there are about
 71,000 large-quantity generators of hazardous waste in the United States.2
 These  generators  account for 99 percent of the 275 million Mg/yr of hazard-
 ous waste produced and managed under RCRA in 1985.3  Hazardous waste gener-
 ators  are most prevalent in the manufacturing industries (standard indus-
 trial  classification [SIC] codes 20-39).  Manufacturing as a whole accounts
 for more  than 90  percent of the total quantity of hazardous waste gener-
 ated.  Among specific industries, the chemical, petroleum,  metals, electri-
 cal equipment, and transportation industries are the major generators of
 hazardous wastes.  Two industry groups that stand out as generators are the
 chemical  and petroleum industries (SIC 28 and 29); these industries alone
 account for more  than 70 percent of total waste generation.  The chemical
 industry  (SIC 28), with only 17 percent of the generators,  generated
 68 percent of all the hazardous wastes produced in 1981.  Another prominent
 group  in the manufacturing sector was metal-related industries (SIC 33-37);
 these  industries  generated about 22 percent of all hazardous wastes in
 1981.4
     The 1981 Survey of Hazardous Waste Generators and Treatment, Storage,
 and Disposal Facilities (Westat Survey)^ showed that only 15 percent of the
 generators were nonmanufacturing or unclassified under SIC.  The survey
 results also provide estimates of number of generators producing specific
types of hazardous wastes.  Just over half the generators indicated that
they generate spent solvents,  both halogenated and nonhalogenated (RCRA
waste codes F001-F005).  Generators of sludges from wastewater treatment
systems associated with electroplating and coating operations £nd gener-
ators of quenching and plating bath solutions and sludges accounted for
16 percent of the generator population.  Only 10 percent of the generators
                                    3-4

-------
generated  listed hazardous wastes from specific  industrial sources  (e.g.,
slop oil emulsion solids from the petroleum refining  industry--K049).
Forty-three percent of generators produce  ignitible wastes (RCRA waste code
D001),  a third generated corrosive wastes  (D002), and more than a quarter
generated wastes that failed EPA's test for toxicity  (D004-D017).   Just
under 30 percent of the generators reported hazardous wastes that were
spilled, discarded, or off-specification commercial chemical products or
manufacturing chemical intermediates  ("P"  and  "U" prefix waste codes).
     The physical characteristics of  the 275 million Mg of RCRA hazardous
waste managed in 1985 vary from dilute wastewater to metal-bearing  sludges
to soils contaminated with polychlorinated biphenyl (PCB).  Over 90  percent
(by weight) of RCRA hazardous waste is in  the  form of dilute aqueous waste.
The remaining wastes are organic and  inorganic sludges and organic  and
inorganic solids.  Figure 3-2 categorizes  hazardous waste by physical char-
acteristics.
     Although small-quantity gene-rators (those that generate more than
100 kg  and less than 1,000 kg of hazardous waste per month) represent a
large proportion of the number of hazardous waste generators nationally
(more than 26,000),^ they account for only a very small fraction of  the
hazardous wastes generated.  About 25 percent  of the country's hazardous
waste generators are small-quantity generators,  but these -generators con-
tribute less than one-half of 1 percent of the total hazardous waste gener-
ated. 8  The majority of the small-quantity generators are automotive repair
firms,  construction firms, dry cleaners, photographic processors,  and
laboratories.  The wastes produced by smal1-quantity generators span the
full spectrum of RCRA hazardous wastes.  According to EPA's National Small
Quantity Hazardous Waste Generator Survey,9 the  majority of smal1-quantity
generator waste is derived from lead  acid  batteries; the remainder  includes
such hazardous wastes as acids, solvents,   photographic wastes, and  dry
cleaning residues.
3.1.3  Transporters
     Once a RCRA hazardous waste is generated, it must be managed (i.e.,
stored,  treated,  or disposed of) in accordance with legal requirements.
Although nearly all  hazardous waste is managed to some degree at the site
                                    3-5

-------
Aqueous Liquids
(252 X  106 Mg)
                                                               Organic Sludges (2 X 106 Mg)
                                                                    Organic Liquids
                                                                    (4X 106 Mg)
                                                                      Inorganic Solids
                                                                      (2X 106 Mg)
                                                                      Aqueous Sludges
                                                                      (15 X 106 Mg)
          Figure 3-2.  Estimate of physical characteristics of RCRA hazardous wastes.6
                                       3-6

-------
where It  is generated, the Westat Survey has shown that only about one in
six generators manage their hazardous waste exclusively onsite.10  of those
generators that  ship hazardous wastes to offsite management facilities for
treatment, storage, and disposal, roughly a quarter still manage part of
their hazardous  wastes onsite.  Although the survey estimated that 84 per-
cent of the generators ship some or all of their hazardous wastes offsite,
the vast  majority of the quantities of hazardous waste are nonetheless
managed onsite.  Data supplied by generators indicate that about 96 percent
of all generated hazardous wastes are managed onsite, with only 4 percent
being shipped offsite for treatment, storage, or disposal.
     In response to the movement of hazardous waste, a large industry has
developed that transports hazardous wastes from generators to TSDF.  It has
been estimated that over 13,000 transporters are involved in moving hazard-
ous wastes by land or water from generators to TSDF.H
3.1.4  Treatment, Storage, and Disposal Facilities
     A significant segment of the hazardous waste industry is involved in
hazardous waste  management (i.e., treatment, storage, and disposal activi-
ties).  Table 3-1 provides the RCRA definition of treatment, storage, and
disposal.  TSDF  must apply for and receive a permit to operate under RCRA
Subtitle  C regulations.  The RCRA Subtitle C permit program regulates 13
categories of waste management processes.  There are four process categor-
ies each  within  storage and treatment practices and five categories within
disposal  practices.  Table 3-2 presents the 13 major categories by RCRA
process code.
     Some of the 13 RCRA process categories can be further classified by
characteristics  of the waste management processes.  For example,  tank
treatment may be quiescent or agitated/aerated (referring to the presence
or lack of movement/mixing of the liquid contained in the tank).   Such
process varieties and similarities are reflected in the characterization of
the industry when estimating nationwide TSDF emissions.  Figures 3-3
through 3-5 provide a more detailed look at examples of the various manage-
ment processes.   As can be seen.from the range of treatment and disposal
processes, the industry is complex and not easily characterized.   The
hazardous waste industry is also dynamic; that is, in response to
                                    3-7

-------
         TABLE 3-1.   RESOURCE CONSERVATION AND RECOVERY ACT (RCRA)
                  HAZARDOUS WASTE MANAGEMENT DEFINITIONS3
      Term
              Definition
Storage
Treatment
Disposal  facility
"Storage" means the holding of hazardous waste
for a temporary period,  at the end of which the
hazardous waste is treated, disposed of, or
stored elsewhere.

"Treatment" means  any method,  technique, or
process,  including neutralization, designed to
change the physical,  chemical, or biological
character or composition of any hazardous waste
so as to neutralize such waste, or so as to
recover energy or  material resources from the
waste, or so as to render such waste non-
hazardous, or less hazardous;  safer to
transport, store,  or dispose of; or amenable
for recovery, amenable for storage, or reduced
in volume.

"Disposal facility" means a facility or part of
a facility at which hazardous  waste is
intentionally placed into or on any land or
water, and at which waste will remain after
closure.
aDefinitions are presented as  stated in RCRA regulations (40 CFR 260.10)
 as of July 1,  1986.12
                                    3-8

-------
       TABLE 3-2.  NATIONWIDE QUANTITY OF HAZARDOUS WASTE MANAGED BY
                            SPECIFIC PROCESSES
       Waste
    management
      process
                   RCRA       Number of active
                  process      facilities with
                   code           process3
                                 Waste quantity
                             managed,3 10^ Mg/yr
Storage
                                                          133
    Container
    Tank
    Wastepile
    Impoundment
Treatment
    Tank
    Impoundment
    Incineration
    Otherb
Disposal
              wel 1
Injection
Landfill
Land treatment
Ocean disposal
Impoundment

Totalc
                    SOI
                    S02
                    S03
                    S04
                    T01
                    T02
                    T03
                    T04
D79
D80
D81
D82
D83
                1,440
                  911
                   57
                  223
                  291
                  127
                  158
                  319
    61
    90
    54
    NA
    47

>2,300
                                                          154
                                                           49
                                                              275
RCRA = Resource Conservation and Recovery Act.
  NA = Not available.

3Based on the 1986 Screener Survey.13  Excludes facilities that manage.
 less than 0.01 Mg/yr in storage, treatment, and disposal processes.
 Quantities were not reported in this survey by specific management
 process.

b"0ther" refers to physical, chemical, thermal, or biological treatment
 processes not occurring in tanks,  surface impoundments, or incinerators.

cFacilities do not add up to about 2,300 because some facilities have more
 than one process.  Waste quantities presented do not add to the total of
 275 million Mg of hazardous waste produced and managed in 1985 because
 some facilities may process a waste in more than one management process.
 For example,  a waste may be stored prior to treatment or treated prior to
 disposal.
                                    3-9

-------
                    Waste
                  Generation
                                                                                  Land Treatment
                                        A.  Land treatment of bulk sludges or liquids
OJ

I—"
o
                          Waste
                       Generation
                                            B.  Land treatment of filtered solids
                       Figure 3-3. Two examples of onsite hazardous waste land treatment operations.

-------
Drums


Inspection
and
Sampling





Free Liquids
Decanted; Drum
Contents Treated
& Fixed in Drum




Free Liquid
Storage
(tank)

Solids

Fixed Waste in Drums


Landfill

j


— &•

Leachate
Storage and
Treatment
Leachate
                                                         Pre-Fixed Waste in Drums




                                                                       A. Drum disposal
HoO
CO
Free Liquids
Tank Trucks
Free Liquids
Pipeline
Sampling
from Trucks
i
r
Liquid Storage
(pond or tank)
i

Product
Recovery
Truck
Pip
Treatment
(tank, pond, or
process vessel)
j
eline
Pipeline


Pipe-
line
Fixation
(open pit)

Truck
Lea
Pif
Landfill
chate
leline ,
j

Solids
Truck
Leachate
Storage and
Treatment
H2O

                                   Sale or Use
                                                                            B. Bulk liquid disposal
                                                      Figure 34.  Two examples of active landfill operations.

-------



Drum
Loading




Fork
Lift 1

w

Drum
Storage



Fork
Lift II


— ^-
OJ
I
                                                                                                        Drum Truck
                                                                                                                       Offsite
                Waste

             Generation
                                          Pump/pipe
Pump/pipe
                       Tank Truck
Offsite
                                                                                                       Dumpster Truck
                                                                                                                         Offsite
                                           Figure 3-5.  Example onsite hazardous waste storage facility.

-------
changing demands  and  regulations, the facilities change the ways wastes are
treated, stored,  and  disposed of.
     The total estimated quantity of hazardous wastes managed at more than
2,300 TSDF  in  1985 was 275 million Mg.  The waste quantities handled by
each of the three main waste management processes (i.e., treatment, stor-
age, and disposal) are presented in Table 3-2.  The waste quantities given
in Table 3-2 will not sum to the total national estimate because some
wastes pass through more than one process; for example, a waste may be
stored prior to treatment or treated prior to disposal.  Also provided in
Table 3-2 is a breakdown of the number of active TSDF by specific type of
treatment,  storage, or disposal process.  In the storage category, con-
tainer storage is a management process utilized by more than half the TSDF;
tank storage occurs at slightly more than a third of the TSDF.  Of the
treatment processes,  tank treatment is widely practiced, but no single
treatment process is  used in a majority of facilities.  In the disposal
category, landfills are the dominant disposal units operated at TSDF.
     The information  presented above is taken from a TSDF data base of
waste management  practices compiled for use in examining the industry and
its environmental and health impacts.  Three data bases were used to gener-
ate this TSDF  data base.  Two major sources were the Hazardous Waste Data
Management  System (HWDMS)^ and the 1981 Westat Survey, both of which are
EPA Office  of  Solid Waste (OSW) data bases.  More recent information from
the OSW 1986 National Screening Survey of Hazardous Waste Treatment, Stor-
age, Disposal, and Recycling Facilities (1986 Screener) was also used to
make the TSDF data base as current as possible.15  Each of these three data
bases provided a different level of detail regarding particular aspects of
the TSDF industry.  For example, the HWDMS provided waste management proc-
ess codes,  wastes codes, and facility SIC codes.  The 1986 Screener pro-
vided information on total annual waste quantities managed by the facility
and operating status  (active or closed)  for the entire industry.  The
Westat Survey,  on the other hand, deals with only a subset of the industry
but provides a greater level  of detail regarding individual facility
operations;  for example,  the distribution of waste quantities handled by
each waste management process is available for each facility in the data
base.
                                   3-13

-------
3.1.5  TSDF Emission Sources
     The organic emission sources associated with each type of storage,
treatment, and disposal process are summarized in Table 3-3.  The emission
sources in this table are arranged into seven categories based on their
common emission characteristics and/or their routine association with other
processes.  These are (1) impoundments and tanks, (2) land treatment,
(3)  landfills and wastepiles, (4) transfer and handling operations,  (5)
injection wells, (6) incinerators, and (7) organic compound removal
devices.
     For open (or uncovered) surface impoundments and tanks, the major
source of organic emissions is the uncovered liquid surface exposed to the
air.  The conditions under which  liquids are stored in uncovered impound-
ments and uncovered tanks ranges from quiescent to highly turbulent since,
in some cases, aeration and/or agitation are applied to aid in treatment of
the waste.  Emissions tend to increase with an increase in surface turbu-
lence because of enhanced mass transfer between the liquid and air.  For
both uncovered and covered storage tanks,  loading and breathing losses are
a major source of emissions.
     At land treatment facilities, wastes  are either spread on or injected
into the soil, after which they are normally tilled into the soil.  Other
activities that are likely to occur at land treatment facilities include
transfer,  storage,  handling, and dewatering of the wastes to be land-
treated.  Examples would include  loading and unloading of wastes in vacuum
trucks and dewatering of wastes using one  of the various types of available
filtration devices.   Each of the  land treatment process stages illustrated
in Figure 3-3 is a potential source of organic air emissions.  The major
emission source associated with land treatment is the land treatment area
itself.
     A landfill  is  a facility,  usually an  excavated, lined pit or trench,
into which wastes are placed for disposal.  Some existing landfills may not
be lined;  however,  all  new facilities are  lined to meet RCRA permit
requirements.   All  wastes containing liquids and destined for disposal in a
landfill  must be treated or "fixed" to form a nonliquid material.  Land-
fills are  a source of organic emissions from several emission points.
                                   3-14

-------
      TABLE 3-3.  HAZARDOUS WASTE MANAGEMENT PROCESS EMISSION SOURCES
     Management process
     Emission source
Impoundments and Tanks

(S04,  T02,  D83) (S02,  T01)

   Quiescent impoundments
     (storage & treatment)

   Quiescent tanks
     (storage & treatment)

       Uncovered
       Covered

   Aerated/agitated impoundments
     (treatment)

   Aerated/agitated uncovered tanks
     (treatment)

   Impoundment lining
   Impoundment inlet
Quiescent liquid surface
Quiescent liquid surface
Working and breathing lossses

Turbulent liquid surface
Turbulent liquid surface
Dredging (exposed waste
  surface)3

Splash loading^
Land Treatment (D81)

   Land application



   Dewatering devices
Land'fills (D80)  and Wastepiles (S03)

   Active landfill
Application of waste to soil
Applied waste before tilling
Applied waste after tilling

Vacuum pump exhaust for vacuum
  f i Hers3
Exposed waste surface in belt
  f i Her presses
Filter cake collection and
  disposal
Transport of waste to landfill
  (open trucks)
Unloading and spreading of
  wastes
Landfilled waste
Leachate (within the confines
  of the 1iner system)

                   (continued)
                                   3-15

-------
                           TABLE  3-3  (continued)
     Management process                           Emission  source  	

Landfills and Wastepiles  (con.)

   Closed landfill                           Landfill  surface  gas  vents  and
                                              manholes
                                            Leachate  (within  the  confines
                                              of  the  liner system)

   Wastepiles                               Wastepile surface
                                            Leachate  (within  the  confines
                                              of  the  liner system)

   Waste fixation

     Pit and mixer                          Splash  loadinq into fixation
                                              pitsb
                                            Mixing  of waste and fixative
                                            Mechanical  mixer  vents

     Drum                                   Drum  inspection3
                                            Drum  decanting3
                                            In-drum fixation3

Transfer and Handling Operations  (SOI,  S02)

   Vacuum trucks                            Vacuum  pump exhaust
                                            Spills  during  truck loading
                                            Truck cleaning3

   Open dump trucks                          Waste surface  during  loading
                                              and transport
                                            Spills
                                            Truck cleaning3

   Equipment leaksc                          Losses  from pumps,  valves,
                                              sampling  connections,  open-
                                              end lines,  and  pressure-
                                              re! ief  devices
   Containers
     Drums
     Tank trucks
     Railroad  tank  cars
     Marine  tankers
     Barges
     Dumpsters
Waste loading
Spillage in transit
Spillage during waste loading/
  unloading
Exposed waste surface
Cleaning losses3

                   (continued)
                                   3-16

-------
                           TABLE 3-3 (continued)
     Management process
       source category                           Emission source

Injection Wells (D79)d

Incinerators (T03)e                         Exhaust gas stacks

Organic Compound Removal (Treatment)"^       Process vents
Devices                                     Condenser vents
                                            Equipment leaks (pumps,
                                              valves, etc.)


aNo emission estimating method exists for this source.

^No emission estimating method exists for this source.  Unlike enclosed
 sources such as tanks, this is an open source and vapor saturation does
 not occur.

cEmissions from equipment leaks are associated with all management
 processes that involve the use of pumps, valves, sampling connections,
 open-ended lines, and pressure-relief devices.

djhis management process is being regulated under a different standard.
 The equipment leak emissions related to the injection well disposal
 process are evaluated in this document.

elncinerator emission sources, such as exhaust gas stacks, are regulated
 under 40 CFR 264, Subpart 0, "Incinerators."  The equipment leak emissions
 related to incineration are evaluated in this document.

fThese devices are chemical process units that are designed to reduce the
 organic content of a waste.  Typically, these processes are employed for
 the recovery of valuable organics for recycle and reuse; in relation to
 environmental protection,  these processes are or can be used for pollution
 control.  For example, organics may be removed to make a waste suitable
 for hazardous waste land disposal.
                                   3-17

-------
 Figure 3-4 shows typical process stages for two variations in landfill
 processing; each of the processing steps identified is a potential emission
 source.  The landfill surface, whether open, covered with earth daily, or
 closed with a cap is an emission source.  A waste fixation pit is another
 source of organic emissions that could be associated with landfills.  Acti-
 vities at the landfill, such as waste transport and waste unloading and
 spreading, are also sources of emissions.  Wastepiles are similar to  land-
 fills and the same emission sources can be found; they are,  in essence,
 temporary landfills.
     Each of the process steps illustrated in Figure 3-5 is a potential
 emission source associated with hazardous waste transfer, storage, and
 handling operations.  Loading operations contribute to overall emissions,
 especially splash loading of waste as opposed to submerged loading.   Spills
 also occur during waste transfer and handling and,  for liquid wastes  that
 are pumped, emissions may occur from fugitive sources such as pumps and
 valves or at open-ended lines, pressure-relief valves, and sampling connec-
 tions.  Organic emissions are associated with all three of the storage
 methods shown in Figure 3-5:  drums, dumpsters, and tanks.
     Miscellaneous sources of emissions such as drum cleaning or the  crush-
 ing and landfilling of empty drums containing waste residues also contrib-
 ute to organic air emissions.  The improper handling of drum residue  can
 lead to emissions along with waste residues lost to the environment by
 uncontained drum crushing operations.  In addition, RCRA permit conditions
 require annual  dredging of surface impoundments; the dredging operation, a
waste transfer process, may also be a source of organic emissions.
3.2  ESTIMATES OF ORGANIC EMISSIONS
     A modeling  approach based on applicable mass transfer equations  was
selected  as the  method of estimating organic emissions from TSDF.  Models
initially developed by the EPA Office of Solid Waste were refined to  incor-
porate inputs  relevant to estimating air emissions.  The models selected
for use are formulated for individual management processes at TSDF and
account for such factors as process design and operating parameters as well
as the meteorological  effects on emissions.  These emission models were
used to generate estimates of the amounts of organics in the incoming
                                   3-18

-------
wastes that are emitted to the air or biodegraded during processing.  More
traditional emission estimating techniques and methodologies, such as
basing emission estimates on the results of a limited number of actual TSDF
source tests, are not appropriate for the diverse operations found in this
industry.  The TSDF industry and its waste management processes are too
varied to use source test data as the sole basis to estimate industry-wide
emissions.
     The use of emissi'on models makes it possible to generate emission
estimates under a wide variety of source conditions and waste compositions.
The accuracy of estimates made by the emission models in comparison to
actual field measurements of emissions for specific sites has been exam-
ined.  That comparison is discussed in Appendix C.  In general, it was
found that, where comparisons could be made, emission model estimates com-
pared favorably with field data.  Appendix F contains the TSDF source test
data.
     The following sections describe the bases for developing estimates of
organic compound emissions from TSDF using the modeling approach.   Section
3.2.1 discusses the elements necessary for producing nationwide emission
estimates for individual waste management units.  Section 3.2.2 discusses
how those elements are combined in a single computer model to produce the
nationwide emission estimates.
3.2.1  Emission Estimation Data Requirements
     Key elements in the estimation of organic emissions from TSDF are the
availability of:  (1) facility-specific information, (2) management process
emission characteristics, and (3) waste compositions.  Facility-specific
information and data include the types of waste management processes pres-
ent in those facilities, the RCRA waste codes managed at the facilities,
and the total quantity of waste managed for each of the facilities nation-
wide.  Some facility-specific information is available through data bases
established for other EPA projects (e.g., the HWDMS, the 1981 Westat
Survey,  and the 1986 Screener).   Where facility-specific data are insuffi-
cient,  estimated values based on typical industry wastes and operating
practices are used.   Section 3.2.2.1 describes the facility information
available for the generation of nationwide estimates.
                                   3-19

-------
      In addition to facility-specific data, emitting characteristics of the
waste management units are needed to estimate nationwide emissions.  Typi-
cally, emission measurements are made at the source and those measurements
serve as the basis for characterization of similar emission sources.  In
the case of TSDF, there is a diversity of sources and factors within a
source that have a significant impact on emissions, waste composition being
a major one; this variation makes estimation of nationwide TSDF emissions
directly from only measured data impractical.  Other factors that restrict
this  approach include the overall lack of emission test data for the range
of TSDF management processes and the absence of standardized test methods
that  allow meaningful comparisons of available emission data to be made.
As an alternative, emission models have been adapted to facilitate genera-
tion  of waste management process emission estimates.  These emission models
are presented in Appendix C, Section C.I.  To use the emission models, it
is necessary to define certain waste management unit design and operating
characteristics (such as surface area and waste retention time for surface
impoundments).  Given that this level of detail is not available for most
facilities, process parameters based on model management units were devel-
oped  for use in calculating emissions (also, costs of control and emission
reductions).  Using survey results and information from other sources such
as design manuals and site visit reports, model units were developed in
terms of operating and design parameters spanning typical ranges of surface
area,  retention times, and other characteristics representative of the TSDF
industry.   A sensitivity analysis was conducted for each model to determine
which input parameters,  over what range, have significant effects on emis-
sion model  estimates.  Appendix C,  Section C.2, discusses the sources of
information and rationale used to develop the TSDF model units and lists
the specific characteristics of each model waste management unit that are
needed to  compute emissions using the appropriate emission model.
     The other key element in estimating emissions is the composition of
the waste  in the waste management unit.  The specific chemicals found in
each waste  management unit nationwide are not known, but data are available
on waste composition from a number of facilities in several industrial
categories.  These data have been combined into a data base that gives
                                   3-20

-------
waste composition as a function of SIC and RCRA waste codes.  The file is
described briefly in Section 3.2.2.2 and described in detail in Appendix D,
Section D.2.2.
     Table 3-4 presents the relative emissions predicted by the emission
models for selected model waste management units.  This table lists pre-
dicted uncontrolled organic emissions by model unit for five different
model waste compositions.  The specific compositions of the five model
wastes are given in Appendix C, Section C.2.2, along with the rationale for
their development.  The results in Table 3-4  illustrate the variability in
emissions that may occur from waste management units for different waste
compositions.  The table also shows emission  variability between waste
management units for the same waste type.  No conclusions should be drawn
from this latter comparison without considering the differences in waste
throughput between the waste management units.  It should be pointed out
that, to the extent possible, the composition and quantities of the actual
waste streams processed at the existing facilities were used in estimating
nationwide emissions.  The model wastes are presented here to illustrate
the variability in potential air emissions in relation to waste composition
and management process.
     In calculating nationwide TSDF air emissions, emission models are used
with the model waste management unit design and operating characteristics
to produce emission factors for the model units.  The model unit emission
factors are estimates of the fraction of specific organic compounds enter-
ing the waste management unit that become air emissions from that unit.
Derivation of these emission factors involves combining the steps discussed
previously in this section with a knowledge of the properties of the com-
pounds for which emi-ssion factors are required.  The development of emis-
sion factors is explained in detail in Appendix D, Section D.2.4.
3.2.2  Nationwide TSDF Emissions
     Nationwide organic emissions from the TSDF industry were estimated
using the Source Assessment Model  (SAM),  a computerized simulation program
designed to generate nationwide emissions estimates on a facility, waste
management unit,  or emission source basis.  Summation of individual facil-
ity results provides the nationwide emission  estimate.  The SAM utilizes a
                                   3-21

-------
TABLE 3-4.   SUMMARY OF  SELECTED  MODEL  HAZARDOUS  WASTE MANAGEMENT  UNIT
  UNCONTROLLED ORGANIC  EMISSION  ESTIMATES  FOR  MODEL  WASTES  (Mg/yr)a
Model unitb
Covered storage tank
(S02D)
Covered quiescent
treatment tank (T01E)
Quiescent uncovered
storage tank (S02I)
Quiescent uncovered
treatment tank (T01B)
Quiescent storage
impoundment (S04C)
Quiescent treatment
impoundment (T02D)
Quiescent disposal
impoundment (D83A)
Uncovered aerated/
agitated treatment
tank (T01G)
Aerated/agitated treat-
ment impoundment
(T02J)
Waste fixation (Fixation
Pit B)
Drum storage (S01B)
Dumpster storage (S01C)
Wastepiles (S03E)
Model waste type
Aqueous Organic
sludge/ Dilute Organic sludge/
slurry aqueous , liquid slurry
0.117 2.12 0.437 1.11
0.24 4.6 1.19 1.40
24 8.1 514 586
34 19 954 1,026
686 159
946 269
842 130
870 130

1,920 390
4,110
0.0036 0.0000909 0.0236 0.0298
0.72 -- 0.049 1.44
139.7

Two-phase
aqueous/
organic
0.891
1.94
9.7
31
183
326
16,000
..

380
31,700
0.000181
--
100
                                                              (continued)
                               3-22

-------
                            TABLE 3-4 (continued)
Model unit°
Landfi 1 1 -active
Landfi 1 1-closed
Land treatment

(D80E)
(D80H)
(D81C)

Aqueous
sludge/
slurry
358
0.068
269
Model waste type
Organic
Dilute Organic sludge/
aqueous liquid slurry
__
__
21.6

Two-phase
aqueous/
organic
299
2.09
--
-- = Indicates this model  unit does not manage this model  waste type,  thus an
     uncontrolled emission estimate is not available.

aThis table lists the estimated organic emissions for selected model  waste
 management units when the listed model wastes are managed in those units.
 The model  unit definitions are given in Appendix C,  Section C.2.1.  The
 model  waste compositions  are also described in Appendix C,  Section C.2.2.
     parenthetical  listings are the model  unit designations under which the
 model  unit definitions can be found in Appendix C,  Section C.2.2.
                                    3-23

-------
 variety of  information and data concerning the TSDF industry to calculate
 emissions.  The SAM processes the information and data from a number of
 input  files that contain TSDF-specific information (facility location,
 waste  management processes used, and types and quantities of wastes man-
 aged), waste characterization data (approximate compositions of typical
 wastes), and air emission model estimates (emission factors based on char-
 acteristics of both TSDF waste management units and waste types).
     Because of the complexity of the TSDF industry and the current lack of
 detailed information for all TSDF, it is unlikely that the SAM estimates
 are  accurate for an individual facility.  However, it is believed that the
 SAM  emission estimates are a reasonable approximation on a nationwide basis
 and  the TSDF modeling approach provides the best basis for analysis of
 options for controlling TSDF air emissions.
     A brief discussion of the input data files, assembled for and used by
 the  SAM to calculate air emissions,  and the output emissions files gen-
 erated by the SAM are presented in the following sections of this chapter.
 Figure 3-6 outlines the main SAM files and functions used in estimating
 nationwide emissions from the TSDF industry.   The SAM, its data inputs and
 outputs, and the overall logic used in the model's calculations are dis-
 cussed in more detail  in Appendix D.
     3.2.2.1  SAM Input Files.  There are four main data files that are
 inputs to the SAM .nationwide uncontrolled emission estimates:  the Industry
 Profile (a file of waste management practices for each TSDF in the Nation),
 the waste characterization file (also referred to as the Waste Characteri-
 zation Data Base),  the chemical properties file, and the emission factors
 file.  These inputs provide the information and data necessary to calculate
 nationwide TSDF uncontrolled emissions.
     3.2.2.1.1  Industry Profile.  The Industry Profile data base was
 developed to provide a list of TSDF nationwide and to describe facility-
 specific waste management practices in terms  of the types and quantities of
wastes handled and the processes utilized.  Several hazardous waste indus-
try surveys and data bases, available through EPA's Office of Solid Waste,
serve as the basis  of  the SAM Industry Profile (see Appendix D, Section
D.2.1).  The information and data from each of these surveys and data bases
have been adapted to fit the needs of the SAM.
                                   3-24

-------
 INPUT FILES
                           OUTPUT
   Industry
    Profile
    Waste
Characterization
      File
   Chemical
  Properties
      File
   Emission
    Factors
     File
  SAM
Processor
Nationwide TSDF
  Uncontrolled
   Emissions
    Figure 3-6.  Source Assessment Model (SAM) input files used in estimating nationwide
        treatment, storage, and disposal facilities (TSDF) uncontrolled air emissions.
                                      3-25

-------
     The information that the SAM uses from the Industry Profile to esti-
mate nationwide emissions includes the following for each TSDF:  (1) facil-
ity identification number (FCID), (2) location coordinates of the facility,
(3) the primary SIC code for the facility, (4) the RCRA waste codes managed
at the TSDF,  (5) the waste quantity for each of the waste codes, and
(6) the management process codes applicable to each waste code.  It is
important to  note that the SIC and waste codes link the facility to the
waste characterization file, which gives estimated waste compositions.
     3.2.2.1.2  Waste characterization file.  This waste characterization
file contains waste data that have been compiled to represent chemical -
specific waste compositions for each waste found within an SIC code.  An
RCRA waste may be generated in one of several  physical/chemical forms; for
example, a waste may be an aqueous liquid or an organic sludge.  The waste
characterization file contains the composition of waste streams in terms of
chemical constituents and their respective concentrations for each physi-
cal/chemical  form of a waste associated with a particular RCRA waste code
in an SIC category.  If specific chemical constituents were not found in
the original  data, chemical  assignments were made based on a review of
similar TSDF  processes.  Wherever available, specific chemicals were
retained in the waste characterization file.  The data provided in the
waste characterization file are accessed by the SAM for each TSDF emission
calculation.  (See Appendix D, Section D.2.2,  for a more detailed discus-
sion.)
     3.2.2.1.3  Chemical  properties file.  Emission estimation for each of
the more than 4,000 waste chemical constituents identified in the waste
characterization file would require property data for all compounds; many
of which are not available.   Therefore, to provide the emission models with
appropriate constituent physical, chemical, and biological properties, the
waste constituents were categorized and grouped into classes based on
volatility (i.e.,  vapor pressure or Henry's law constant) and
biodegradation.   These categories were defined to represent the actual
organic compounds  that occur in hazardous waste streams and serve as
surrogates for the particular waste constituents in terms of physical,
chemical,  and biological  properties in the emission calculations carried
                                   3-26

-------
out by the SAM.   (See Appendix D, Section D.2.3, for a more detailed
discussion.)
     3.2.2.1.4  Emission factors file.  For each waste management process
(e.g., aerated surface  impoundment or treatment tank), the respective emis-
sion models applicable  to the process were used to determine the amount or
fraction of the organic compound entering the unit that is emitted to the
air and the fraction that is biodegraded.  The calculations were made for
each chemical surrogate category for each waste management process.  In
addition to emission factors for process-related emissions, emission fac-
tors developed for transfer and handling-related emissions were also incor-
porated into the  SAM program file.  The emission factors used for estimat-
ing TSDF emissions in this document were calculated using the TSDF air
emission models as presented in the March 1987 draft of the Hazardous Haste
Treatment, Storage, and Disposal Facilities:  Air Emission Models, Draft
Report.16  Since  that time, certain TSDF emission models have been revised
and a new, final  edition of the air emission models report has been
released (December 1987).^  The principal changes to the emission models
involved refining the biodegradation component of the models to more accu-
rately reflect biologically active systems handling low organic concen-
tration waste streams.  With regard to emission model outputs, the changes
from the March draft to the December final version affect, for the most
part, only aerated surface impoundments and result in a minor increase in
the fraction emitted for the chemical surrogates in the high biodegradation
categories.18  For the  other air emission, models, such as the land treat-
ment model, which were  also revised to incorporate new biodegradation rate
data, the changes did not result in appreciable differences in the emission
estimates.   (Appendix D, Section D.2.4, contains a more extensive discus-
sion of emission factors.)
     3.2.2.2  Uncontrolled Nationwide Emissions.  The SAM computes nation-
wide uncontrolled TSDF emissions by first identifying particular waste
management process units within the facility from the Industry Profile.
Once a management process is identified,  the SAM then calculates emissions
on a chemical-by-chemical basis.  The quantity of a particular chemical in
the waste stream is multiplied by the appropriate emission factor, which  is
determined by the chemical,  physical, and biological properties of the
                                   3-27

-------
chemical.  Emissions for the unit are the sum of the emissions for each
chemical constituent in the waste stream.  Emissions for each management
process unit can then be summed; emissions from source categories (manage-
ment units with similar emission characteristics,  e.g., quiescent storage
impoundments and quiescent treatment impoundments)  are then summed to yield
a nationwide emission estimate.
     The nationwide emission estimates for the current TSDF community are
based on 1985 data containing general operating conditions and practices,
the time covered by the most recent TSDF industry  survey.  These emission
estimates are considered to represent the uncontrolled situation or case;
review of the existing applicable State regulations has shown a wide varia-
tion in level of control required for these sources,  with many States
having no control  requirements for TSDF.
     The uncontrolled nationwide TSDF emission estimate as determined by
the SAM is 1.8 million Mg of organic emissions annually.   The breakdown of
nationwide emissions by source category is provided in Table 3-5.  (Chap-
ter 6.0 presents additional  information on these uncontrolled emissions.)
Table 3-5 shows that storage tanks are estimated to be the single largest
emitting source nationwide.   Treatment tanks  and impoundments that are
aerated to promote biological  activity are the second highest single
source.  These two source categories combined account for about 70 percent
of the annual emissions estimated.
                                   3-28

-------
   TABLE 3-5.  NATIONWIDE UNCONTROLLED TSDF ORGANIC EMISSION ESTIMATES3

                                      Nationwide uncontrolled emissions,
Source category
Drum storage
Dumpster storage
Storage tanks
Quiescent surface impoundments^
Quiescent treatment tanks
Aerated/agitated tank and surface
impoundments
Wastepi les
Landfills
Waste fixation
Incineration0
Land treatment
Other treatment^
Spills
Loading
Equipment leaks
Total
1(P Mg/yr
0.19
78
756
209
48
515

0.13
40
2.1
0.89
73
d
0.43
6.8
80
1,810
TSDF = Treatment, storage, and disposal  facility.

aThis table presents the nationwide estimates of uncontrolled TSDF organic
 emissions generated by the Source Assessment Model  described in  Appendix
 D.  Emissions are presented for management processes that have similar
 emission characteristics, i.e., source categories.
^Includes quiescent surface impoundments used for  both storage,  treatment,
 or disposal .

cUncontrol led  incinerator emissions includes emissions from wastes that
 are routinely incinerated with stack exhaust gas  emission controls.
 These sources are currently regulated under 40 CFR  264 Subpart 0.  The
 uncontrolled  emission scenario does not include wastes that are  or would
 be incinerated as a result of implementing the RCRA land disposal
 restrictions  (LDR).  The baseline and two example control strategies do,
 however,  account for the incinerator emissions resulting from the LDR.
 The emission  scenarios are explained in Chapter 5.0.
     r treatment includes processes such as stream stripping that are
 typically used to remove organics from wastes.   For the uncontrolled
 emission case, these emissions are built into the tank treatment
 category due to similarities in emission characteristics.


                                   3-29

-------
3.3  REFERENCES

1.   Memorandum from Coy, David, RTI, to D.ocket.  December 9, 1987.
     Hazardous waste treatment, storage, and disposal facility (TSDF) uni-
     verse of waste constituents.

2.   U.S. Environmental Protection Agency.  Summary Report on RCRA Activi-
     ties for May 1986.  Office of Solid Waste.  Washington, DC.  June 16,
     1986.  p. 4.

3.   U.S. Environmental Protection Agency.  The Hazardous Waste System.
     Office of Solid Waste and Emergency Response.  Washington, DC.  June
     1987.  p. 1-4.

4.   Westat, Inc.  National Survey of Hazardous Waste Generators and Treat-
     ment, Storage and Disposal Facilities Regulated Under RCRA in 1981.
     Prepared for the U.S. Environmental Protection Agency, Office of Solid
     Waste.  April 1984.  p. 141.

5.   Reference 4, p. 65.

6.   Reference 3, p. 2-3.

7.   Reference 2, p. 4.

8.   Abt Associates, Inc.  National Small Quantity Hazardous Waste Genera-
     tors Survey.  Prepared for the U.S. Environmental Protection Agency,
     Office of Solid Waste.  Washington, DC.  February 1985.  p. 2.

9.   Reference 8, p. 2.

10.  Reference 4, p. 69.

11.  Reference 2, p. 4.

12.  U.S. Environmental Protection Agency.  Code of Federal Regulations.
     Title 40,  part 260.10.  Washington, DC.  Office of the Federal Regis-
     ter.  July 1, 1986.

13.  Memorandum from Maclntyre, Lisa, RTI, to Docket.  November 4, 1987.
     Data from the 1986 National Screening Survey of the Hazardous Waste
     Treatment,  Storage, Disposal and Recycling Facilities used to develop
     the Industry Profile.

14.  Memorandum from Maclntyre, Lisa, RTI, to Docket.  November 4, 1987.
     Data from the National Hazardous Waste Data Management System used to
     develop the Industry Profile.

15.  Reference 13.
                                   3-30

-------
16.   Research Triangle Institute.   Hazardous Waste Treatment,  Storage and
     Disposal Facilities:   Air Emission Models,  Draft Report.   Prepared for
     U.S.  Environmental  Protection Agency.   Office of Air Quality Planning
     and Standards.   Research Triangle Park, NC.   March 13,  1987.

17.   U.S.  Environmental  Protection Agency.   Hazardous Waste  Treatment,
     Storage, and Disposal  Facilities (TSDF)--Air Emission Models.
     Research Triangle Park,  NC.   Publication No. EPA-450/3-87-026.
     December 1987.

18.   Memorandum from Coy,  David,  RTI, to Docket.   December 18,  1987.   Com-
     parison of weighted average  emission factors.
                                   3-31

-------
                         4.0  CONTROL TECHNOLOGIES

     This chapter identifies and describes technologies for reducing
organic air emissions from hazardous waste treatment, storage, and disposal
facilities (TSDF) for the protection of human health and the environment.
For each control technique identified, applicable wastes and waste manage-
ment processes are identified as are the estimated levels of control that
can be achieved by each technique in each application.  Cross-media and
secondary environmental impacts associated with the use of a control
technique also are identified.
     There are several general types of controls for reducing organic air
emissions, including organic removal and hazardous waste incineration,
emission suppression by capture or containment, the use of add-on emission
control devices, and process modifications and improved work practices.
Table 4-1 lists these types of emission controls and presents several
emission reduction options that correspond to each control type.  If
organics are removed from either the waste stream prior to disposal or from
an effluent stream, they must be managed appropriately to prevent air
emissions or other adverse environmental impacts.  It should also be noted
that emission suppression does not necessarily result in a reduction in
overall emissions.  Unless emission suppression procedures are used in
conjunction with organic removal or incineration processes, suppression may
serve only to shift emissions downstream in a waste management process or
to spread emissions over a longer period of time.  Used alone, emission
suppression is less desirable than alternatives that remove or incinerate
organics.
     The controls described in this chapter were considered in the develop-
ment of control strategies (combinations of controls and TSDF emission
sources)  for TSDF air emissions.  The emission reduction potential of each
                                    4-1

-------
            TABLE 4-1.  POTENTIAL ORGANIC AIR EMISSION REDUCTION
                          OPTIONS FOR TSDF SOURCES
     Emission control
           type
             Emission reduction
                   options
1.  Suppression controls
2.  Add-on emission control
    devices
    Organic removal/
    hazardous waste
    incineration
    Process modification/
    improved work practices
a.  Enclosures
b.  Covers

a.  Carbon adsorption
b.  Combustion of organic vapors
c.  Condensation

a..  Steam stripping
b.  Air stripping
c.  Thin-film evaporation
d.  Batch distillation
e.  Dewatering3
f.  Waste stream incineration

a.  Coking3 of petroleum refinery wastes^
b.  Submerged loading
c.  Subsurface injection3
d.  Daily earth covers
e.  Mechanical mixers
f.  Improved housekeeping
g.  Leak detection and repair to reduce waste
    transfer fugitive emissions
TSDF = Treatment,  storage,  and disposal  facilities.
                                             a potential control option,
                                             use has not been documented
                                         the
aAlthough this technique is considered to be
 emission reduction that may result from its
 adequately.

bCoking of petroleum refinery wastes has not been considered as a control
 option because petroleum refinery waste is expected to be regulated by the
 land disposal restrictions.  (For further explanation, refer to Chapter 5.0
 Section 5.3.2.1.)
                                     4-2

-------
control was evaluated and, where possible, a control efficiency was
assigned for use  in estimating the nationwide impacts of applying various
strategies.  For  some of the potential controls discussed in this chapter,
emission reductions could not be documented; therefore, they were not
explicitly used in the formulation of control strategies and in the estima-
tion of nationwide impacts of control strategies.
     A summary of the controls used  in the development of control
strategies is presented in Section 4.9, following the discussion of poten-
tial controls in  Sections 4.1 through 4.7.  Cross-media impacts created"
from these controls are addressed in Section 4.8.  Discussion of control
strategies is presented in Chapter 5.0, and the health and environmental
impacts of example control strategies are presented in Chapter 6.0.
Chapter 7.0 presents the cost impacts of the example control strategies.
The model developed to estimate nationwide impacts of control strategies
(referred to as the Source Assessment Model) is described in Appendix D.
4.1  SUPPRESSION  CONTROLS
     Suppression  controls reduce emissions by containing the organics with-
in a confined area and preventing further vaporization.  Unless used in
conjunction with  an add-on control device such as a carbon adsorption unit,
these controls may shift the emissions further downstream in the waste
management process or spread the emissions over a longer period of time.
4.1.1  Enclosures
     4.1.1.1  Air-Supported Structures.  An air-supported structure is an
anchored, flexible membrane dome that is supported by maintaining a posi-
tive pressure under the dome relative to the surrounding atmosphere.  The
dome requires a tightly sealed anchoring system so that the positive pres-
sure can be maintained.  Anchoring typically involves attaching the struc-
ture to a concrete footing around the perimeter of the dome.  A large fan,
drawing some outside air,  provides the necessary pressure differential,
usually about 10.1 kPa (0.1 atm),  between the inside and outside of the
structure.   In addition,  an exhaust  vent and an airlock entrance are
provided for air removal/collection  and access,  respectively.  The exhaust
vent allows installation of a pollutant control  system, such as a carbon
adsorption  unit.   An air-supported structure is illustrated in Figure 4-1.
                                    4-3

-------
       VINYL-COATED
      POLYESTER BASE
          FABRIC
BIAS HARNESS
 NET SYSTEM
        INFLATION/HEATING
             SYSTEM
VEHICULAR
AIR-LOCK
                  PERSONNEL DOOR
Sourcat Air  Structures  International,  Inc
            Figure 4-1. Typical air-supported structure.1
                      4-4

-------
     Complete enclosure by an air-supported structure is feasible if vapor
collection and disposal methods that minimize air emissions are employed,
if the dome materials are compatible with the vapors, and if the dome is
built to withstand the local weather.  If high humidity levels are expected
inside the structure and carbon adsorption is used for emission control, it
is important that the system design take into account the tendency of mois-
ture to reduce the adsorptive capacity of carbon beds (see Section 4.2.2).
Air-supported structures can be built to cover large areas and have been
designed to cover areas of 4 acres and greater.2,3  Tney are appropriate
for hazardous waste management processes that are compatible with the
limited access available through the airlock entranceways and with reduced
sunlight and wind.  Treatment and storage impoundments and landfills are
the waste management processes that are most appropriately controlled by
air-supported structures.
     Air-supported structures have been designed and built for both
impoundments and  landfills.  There is one documented installation of an
air-supported structure for odor control.4  Vendors of air-supported struc-
tures have indicated the existence of several installations for emission
control.5  Although other waste management processes such as wastepiles and
other storage units could be controlled by enclosures,  no known installa-
tions had been built as of 1986.  In some cases,  enclosures could be used
to control multiple management processes, but the physical problem of
access could limit such applications.
     The efficiency of air-supported structures in reducing or suppressing
emissions is determined by (1) the reduction in wind- or sunlight-induced
volatilization of organics, and (2)  the combined effects of the capture
efficiency of the structure and the removal efficiency of the control
device in removing organics.  An enclosure alone may not reduce emissions,
but because it allows venting of the exhaust air to a control  device, high
theoretical  emission reductions are achievable.  However,  tests at one
facility showed that actual control  efficiencies were less than theoretical
or predicted levels due to adverse moisture effects on the carbon adsorber
control  device.6  Factors that affect control efficiency include preventing
air loss through sources other than the vent (e.g., seals, doors, ducting)
                                    4-5

-------
and the efficiency of the control device in removing organics from the
exhaust stream.  Because air-supported structures enclose the entire area
of the emission source, the efficiency of the enclosure in capturing
emissions should be essentially 100 percent.  Thus, the overall control
efficiency of an air-supported structure vented to a control device will be
equivalent to the efficiency of the control device.
     Air-supported structures could be vented to several types of control
devices such as thermal or catalytic incinerators and carbon adsorption
systems.  These alternative devices are discussed in Section 4.2.  The
selection of the control device is dependent on exhaust air flow rate,
concentration of organics in the exhaust air, and the specific organic
pollutants present.  Carbon adsorbers were selected for use in generating
estimates of nationwide controlled emissions and costs for vented air-
supported structures on the basis of comparative model unit costs.  How-
ever, the specific site conditions at a TSDF might cause an alternative
device with equivalent performance, such as a vapor incinerator, to be less
costly.  For the purpose of estimating nationwide emissions, an overall
control efficiency of 95 percent is used for air-supported structures
vented to a carbon adsorber, which is based on the control efficiency of
carbon adsorption as discussed in Section 4.2.1.  (See Appendix F, Section
F.I.2.2, for a discussion of an air-supported structure.)
     An air-supported structure that is vented to a carbon adsorber has
associated cross-media and secondary environmental impacts.  Carbon
adsorbers generate cross-media impacts in the form of spent carbon and if
regenerable carbon beds are used, effluent streams of condensed steam and
organics are generated.  Regenerable carbon adsorption also has associated
secondary environmental impacts in the form of air emissions from the com-
bustion of fuel to produce steam used in the regeneration process.
     4.1.1.2  Rigid Structures.  "Rigid structures" refers to permanent
buildings designed to contain emissions from stored waste or waste process-
ing operations.  Rigid structures can be used to control waste storage or
waste processing emissions.   Typically, they would be used to enclose drum
storage or drummed waste fixation areas, wastepiles, and pits or mechanical
mixers for fixation of the bulk liquids and sludge wastes normally handled
                                    4-6

-------
at TSDF.   The  designs  of  these  structures  are  variable  and highly  site-
specific.  Although  all-metal structures frequently  are used, other types
of construction  undoubtedly would  be  suitable  for  some  applications.
     The buildings themselves-,  without  added emission control devices,
serve to limit organic emissions by reducing the windspeed across  the
surfaces of exposed  wastes within  the structure.   Emissions are further
reduced if building  air is evacuated  to an emission  control device.
Evacuation imposes a slight negative  pressure  on the building, thereby
minimizing the loss  of airborne emissions  through  openings, in the
structure.
     Three operations  at  TSDF were identified  as having metal buildings
that were  designed and installed- to limit  the  escape of emissions  to the
atmosphere.  Two of  these enclose  waste fixation pits,7,8  ancj the  third
encloses pugmills  (used for waste  fixation) and two  wastepiles.9   All three
of these buildings are evacuated to scrubber systems designed for  both
particulate and  odor control.
     No data on  the  emission reduction  achievable  by either nonevacuated or
evacuated  buildings  were  found  in  the literature.  However, if the build-
ings are evacuated at  a rate sufficient to maintain  negative pressure at
the enclosure  doors, capture efficiency should approach 100 percent, and
overall emission reductions will be equal  to the efficiency of the control^,
device.
    ~Rigid structures may be vented to  a variety of  control devices
depending  on the types of pollutants  (organics, particulates) in the
exhaust air.   Section 4.2 describes some of the control devices for organic
emission control.  As  in  the case  of  air-supported structures, carbon
adsorption and thermal or catalytic incineration are candidates.   The
choice among these would  depend on site-specific factors such as exhaust
air flow rate, organic concentrations in the exhaust air,  and the  specific
compounds present.
     In generating nationwide estimates of controlled emissions and costs,
rigid structures were assumed to be present at TSDF  where  drum storage is
practiced.   This is  a reasonable assumption because  good operating practice
would lead to isolation of the  stored drums from wind,  rain, and other
                                    4-7

-------
 environmental  factors that might adversely affect the containers.  Also,
 the  rigid  structures were assumed to be vented to carbon adsorption
 systems, with  a control efficiency of 95 percent, as discussed  in Section
 4.2.1.  Carbon adsorption was selected, as opposed to incineration, because
 of the  relatively  low Organic concentrations in the exhaust air.  Organic
 emissions  are  likely to be caused mostly by opening drums for inspection or
 by spills, both of which are intermittent.
      Cross-media and secondary environmental impacts associated with the
 installation of rigi'd structures are all directly attributable  to the emis-
 sion  control devices used to clean the exhaust air.  Where scrubbers are
 used, contaminated blowdown water is generated.  If a baghouse  or carbon
 adsorption unit is installed, provision must be made for disposing of the
 baghouse dust  or spent carbon.  With regenerable carbon adsorbers, spent
 carbon  must be disposed of and contaminated steam and organic condensate
 must  be dealt  with.  Carbon regeneration also results in secondary air
 emissions  produced by the combustion of fuel to generate steam  used for
 regeneration.
      4.1.1.3   Pressurized Tanks.  Pressurized tanks are used in the
 chemicals manufacturing industries to store volatile organic liquids.  The'
 extent  to which they are used in TSDF is unknown.  Given that some wastes
 managed in TSDF are similarly volatile, it is reasonable to assume that
 pressurized tanks could be used as a suppression-type control device in
 TSDF.
      A  pressurized tank operates at pressures greater than atmospheric and
 is equipped with a pressure relief valve to prevent excessive pressure
 buildup.  The two general  classes of pressurized tanks are low  pressure
 (120  to 200 kPa [1.2 to 2.0 atm]) and high pressure (greater than 200 kPa
 [2.0  atm)].
      Pressurized tanks generally are used for storage of organic  liquids
with  high vapor pressures,  and they vary in size and shape depending on the
operating pressure of the tank.   Under routine loading,  unloading, and
storage operations, a high-pressure tank can be considered to be  a closed
system with no estimated organic emissions.  In low-pressure tanks, working
emissions may occur during tank filling when internal  vapor pressure
                                    4-8

-------
exceeds the maximum relief operating pressure.  Low-pressure tanks can be
considered a control alternative for small (less than 76 m^) fixed-roof
tanks that store all types of waste.  The emission reduction achieved
depends on the- setting of the pressure-relief valve and the vapor pressure
and temperature of the stored material.  For a pressure-relief valve
setting of 200 kPa  (2.0 atm), reported emission reductions range from 71 to
100 percent; for a setting of 120 kPa (1.2 atm), reported reductions range
from 22 to 45 percent.10  Pressurized tanks are one of many options
available for controlling tank emissions.  However, they are regarded as
generally more expensive than other control options for tanks that can
achieve similar emission reductions.  Therefore, in estimating nationwide
emissions, other control options were used instead of pressurized tanks;
however, pressurized tanks may be a. viable option under some conditions.
No cross-media or secondary  environmental impacts are associated with the
use of pressurized tanks.
4.1.2  Covers
     Three basic types of covers are used for storage vessels containing
aqueous and/or organic liquids:  fixed roofs,  external floating roofs, and
internal floating roofs.  Both the external and internal floating roofs
employ a platform that sits  (floats) on the surface of the stored liquid.
Figure 4-2 shows an example  of each storage vessel cover type.  Other cover
types include floating synthetic membranes and flexible covers.  There are
no cross-media or secondary  environmental impacts associated with the use
of covers.
     4.1.2.1  Fixed Roof.  Fixed-roof tanks are widely used in TSDF to
store liquid wastes containing organics with a wide range of volatilities.
The extent to which fixed-roof tanks are used for waste treatment is not
known.  Those treatment processes requiring infrequent access to the tank
surface or personnel access  or not requiring large clearances between the
liquid surface and a roof for operating equipment are amenable to placement
in fixed-roof tanks.
     A permanently affixed roof for a tank vessel is a rigid structure that
typically is equipped with a pressure/vacuum vent.  This allows the tank to
be operated at a slight pressure differential  with respect to the outside
in order to control  emissions.  The roof also may contain access ports such
                                    4-9

-------
 Pruoure /vacuum
     V.lT.
ge  Racch
                                                 -Hanhal,
  Manhole
         Noizle (For
       •ubaerged fill
        or drainage)
                   Typical fixed roof  tank
                                                              External  floating roof  tank.
                                                                                    C«aC«r Veac
Hanhole
                                                                                          Vapor Space
                 Contact Deck Type                                      Noncontact Deck Type

                                      Internal floating roof tanks.


                                   Figure 4-2.  Storage tank covers.
                                                 4-10

-------
as a manhole and a gauge hatch.  Roof shapes vary from flat to dome or cone
shaped.  Roof shape may be an  important consideration for some applica-
tions.  For example, a high dome-shaped roof may be needed to cover a tank
with a surface-mounted aerator.  Emissions from fixed-roof tanks may be
caused by  leaks, breathing losses, or working  losses.  A fixed-roof tank
may leak due to the seals at the valve, hatch, and manhole.  Breathing
losses, caused by vapor expansion and contraction, are the result of
changes in temperature and barometric pressure.  Losses due to filling and
emptying are called working losses.  As the liquid level changes, the vapor-
space either is forced out or  draws  in fresh air that then becomes
saturated  and expands, forcing air out of the  tank.
     The emission models described in Appendix C, Section C.I.1.1.1
(emissions from uncovered [open-top] tanks) and Section C.I.1.3.2 (fixed-
roof tank  emissions) were used to estimate emission reductions for fixed
roofs applied to uncovered tanks.  The tank design parameters used to
estimate emission reductions are described in  Section C.2..1.1.  (Refer to
model unit S02I [uncovered storage tank] and T01B [uncovered, quiescent
treatment  tank].)  Waste compositions used to  estimate emission reductions
included five model wastes listed in Table C-5, excluding the organic-
containing solid waste and dilute aqueous-2 and -3 wastes.  In addition,
two alternates to each of the  five Table C-5 model waste compositions were
used to estimate emission reductions.  These two alternate waste composi-
tions contain the same organic compounds, b'ut  at different concentrations
than those listed in Table C-5.  The average of the three emission reduc-
tions for  each waste form was  computed.H<12   Under these conditions, the
emission reductions for fixed  roofs  applied to uncovered tanks ranged from
86.4 to 99.2 percent, depending on waste form  (e.g., dilute aqueous or
organic sludge/slurry).  The stated  range includes both storage and treat-
ment tanks.  The differences between storage and treatment tank emission
reductions for a given waste form were small (less than 3.5 percent).  The
range in emission reductions is attributable to the variations in organic
compounds present in the model wastes and their concentrations.
     On the basis of the above analysis, nationwide emission estimates were
made using emission reductions of 86 to 99 percent for a fixed roof applied
to an uncovered (open-top) tank. 13
                                   4-11

-------
     4.1.2.2   Internal Floating Roof.  Fixed-roof tanks with internal
floating roofs are used in the petroleum refining industry to store  liquids
containing volatile organic compounds.  The extent to which they are used
in TSDF for managing the refining or other industry wastes is unknown.
Providing there are no waste components incompatible with internal floating
roof materials of construction, it is reasonable to assume that internal
floating roofs could be applied to fixed-roof storage tanks to suppress
organic emissions from certain types of wastes.
     An-internal floating roof is an aluminum or steel floating structure,
called  a deck, that controls the escape of vapors from within fixed-roof
tanks.  The fixed-roof tank has vents that may allow volatile organics to
be released during filling.  In tanks with internal floating roofs,  the
deck rises and falls with the  liquid level and either floats directly on
the  liquid surface (contact deck) or rests on pontoons several inches above
the  liquid surface (noncontact deck).14  Rim seals slide up and down the
tank wall as the deck moves with the liquid.  F-ixed-roof tanks that  have
been retrofitted to employ an  internal floating deck typically are fitted
with vertical  columns inside to support the roof.15  The deck restricts
evaporation of organics by maintaining a constant volume of space between
the  liquid surface and the deck.  Some emissions still occur through deck
fittings, nonwelded deck seams, and the space between deck and wall.
     Installation of internal  floating roofs are appropriate for storage
tanks or quiescent treatment tanks in which the tank contents and the deck
seal  are compatible.   Most liquid and aqueous wastes can be stored in tanks
equipped with  internal floating roofs, but corrosive wastes may not  be
compatible with deck  seals.  Internal floating roofs are well-tested in
industrial  settings.16
     The effectiveness of internal floating roofs in reducing emissions
relative to a fixed-roof tank  is a strong function of tank size, annual
turnovers,  and vapor  pressure of the stored liquid.  In addition, the type
of roof and seal  system employed in the floating roof also affects the
emission reductions achievable with internal floating roofs.  One analysis
using .a model  tank having 50 turnovers per year and storing volatile
organic liquids showed emission reductions of 93.4 to 97.3 percent,
depending on roof and seal  type, relative to a fixed-roof tank.17
                                   4-12

-------
     Emission reductions were also estimated by using the emission models
described in Appendix C, Section C.I.1.4.2  (fixed-roof tank emissions) and
Reference 18 (emissions from tanks with  internal floating roofs).  The tank
design parameters used  in this analysis  are those described in Section
C.2.1.1, for model unit S02I.  Waste  compositions assumed are those model
wastes listed in Table  C-5, excluding the organic-containing solid and
dilute aqueous-2 and -3.  Under these conditions, the estimated emission
reductions ranged from  74 to 82 percent.19,20  jne  range in estimated
emission reductions is  attributable to the  variations in compositions and
concentrations of the model wastes by waste form, e.g., dilute aqueous vs.
organic sludge.  Emission reductions  estimated using emission models and
model wastes described  in Appendix C  are less than  those in the first
referenced analysis due to the model wastes having  different vapor
pressures and the tank  sizes and throughputs being  different.21  The model
waste compositions and  model unit design parameters provide a better
representation of the spectrum of operating conditions for TSDF than the
first referenced analysis.  For the purpose of estimating nationwide
emissions, calculated emission reductions ranging from 74 to 82 percent,
depending on waste form, are used as  the control efficiency for internal
floating roofs when placed inside a fixed-roof tank.
     Internal floating  roofs may also be applied to an open-top vertical
tank in conjunction with a fixed roof to suppress the uncovered tank
organic emissions.  For this combination, the emission reductions achiev-
able are a combination  of the reduction  from applications of the fixed roof
to the uncovered tank,  plus application  of  an internal floating roof to a
fixed-roof tank.  The range of emission  reductions  achievable based on
combinations of the fixed roof with an internal floating roof is 96 to
99 percent.   This range of emission reductions, depending on waste form,  is
used to estimate nationwide emissions from  uncovered tanks after applying
an internal  floating roof in combination with a fixed roof.  An alternative
to this combination is  application of an external floating roof to the
uncovered tank,  discussed in the next section.
     4.1.2.3  External  Floating Roof.  Tanks with external floating roofs
are used in  the petroleum refining industry to store liquids containing
                                   4-13

-------
volatile organic compounds.  As with internal floating roofs, the extent to
which external floating roofs are used in TSDF for managing the refining or
other industry wastes is unknown.  Except for incompatibility of the
materials of construction with the stored or treated wastes, it is reason-
able to assume that an external floating roof could be applied to uncovered
vertical tanks to suppress organic emissions from certain types of wastes.
     The external floating roof is similar to an internal floating roof,
except there is no permanent fixed roof on the top of the storage tank.
The deck rises and falls with the surface of the liquid.  The space between
the deck and the vessel wall is controlled by a seal or seal system.  The
effectiveness for controlling organic losses is determined by the ability
of the seal to fill the spaces between deck and wall.  One analysis using a
model tank having 50 turnovers per year showed emission reductions ranging
from 26 to 99 percent depending on seal  type, relative to a fixed-roof
tank.22
     In developing cost and application data for applying external floating
roofs to open-top tanks, equipment vendors and consultants were contacted.
Two vendors indicated that external floating roofs were generally not
applied to tanks of less than 9 meters (30 feet) diameter.23,24  /^n
engineering consultant indicated that maintenance costs for the external
floating roofs were higher than for internal floating roofs.25  This
consultant suggested that the additional  capital cost for an internal
floating roof combined with a fixed roof could be recovered in 3 to 5 years
in reduced maintenance costs as compared to the external floating roof.
     The TSDF statistics used to derive the model unit definitions as
described in Appendix C, Section C.2.1,  suggest that the TSDF storage and
treatment tanks typically tend to be smaller in diameter than 9 meters.  In
fact,  only the largest defined model tank units are about 9 meters in
diameter;  the others are smaller.  This fact, when combined with the vendor
comment indicating relatively high maintenance costs offsetting lower
capital  costs for external  floating roofs, leads to the selection of
internal  floating roofs in combination with a fixed roof as the preferred
type of floating roof for national controlled emissions and control cost
estimates.   This does not mean that external floating roofs are not a good
                                   4-14

-------
choice.  For certain site-specific conditions, an external floating roof
may be a better choice than the combination of internal floating roof
combined with a fixed roof.
     4.1.2.4  Floating Synthetic Membrane.  Floating membranes are
applicable to quiescent impoundments and uncovered storage tanks including
concrete-lined impoundments, which are defined as storage tanks.  They have
been used to cover potable water supplies since the early 1970's and have
been used at Superfund sites to restrict inflow of rainwater.26  They also
have been used as odor-control devices for surface impoundments.27
     Floating synthetic membranes are covers that float on aqueous/liquid
surfaces; they must be made of materials that are (1) chemically resistant,
(2) temperature- and weather-resistant,  (3) relatively impermeable to gases
and vapors, and (4) physically resistant to tearing and stretching.  They
are designed to float directly on a liquid surface or on pontoons.  Pontoon
systems are designed as gas collection systems that vent off-gases to a
control device.28  A floating synthetic membrane that floats directly on
the liquid surface will not, for the most part, allow gases and vapors to
form under the membrane.  Floating membranes are manufactured with cables
for anchoring; concrete or dirt footings are used to keep the membrane in
place.29  Except under the most arid conditions, a rainwater collection
system must be used to remove water that collects on top of the membrane.30
     Emission reductions greater than 90 percent can be expected for com-
patible liner-aqueous waste systems.31  Emission reductions are determined
by the fraction of surface area covered or, in the case of a pontoon sys-
tem, by the control efficiency of gas collection and removal systems, if in
place, and by the permeability of the membrane.
     One synthetic membrane vendor reports that if the membrane floats
directly on the liquid surface and the membrane perimeter is completely
sealed to prevent entrance or exit of air, the only organic loss mechanism
is permeation through the liner.32  However, the vendor also reports that
complete sealing at the perimeter is not typical for reasons such as liquid
level  changes in the impoundments.  Under the condition that a portion of
the impoundment surface area is uncovered, the losses due to permeation are
much smaller than losses from the uncovered surface.
                                   4-15

-------
     An analytical approach was taken to estimate emission reductions for
floating synthetic membranes in the absence of measured data.  Computations
were made for the three model unit impoundment surface areas (model units
are defined in Appendix C, Section C.2.1.1) using a wall  slope of 3:1
(horizontal to vertical) and assumed liquid level changes that would leave
a portion of the liquid surface uncovered.33  This analysis was used to
select 85 percent as the emission reduction basis for floating synthetic
membranes for purposes of estimating nationwide emissions from membrane
covered impoundments.
     4.1.2.5  Flexible Covers.   Flexible covers are synthetic membranes
used to suppress or limit emissions from area organic emission sources.  A
typical cover material is 30-mil high-density polyethylene (HOPE).  This
material is readily available and easily seamed to form sheets large enough
to cover any area desired (HOPE currently is used in landfill liners that
occupy many acres).
     Flexible covers suppress organic air emissions by imposing a vapor
barrier between the emission source and the atmosphere.  They would be most
applicable to landfills and to  wastepiles and other temporary sources of VO
emissions such as loaded open trucks or dumpsters.  In addition to having
suppressed air emissions, covered sources would no longer be exposed to
precipitation that could otherwise infiltrate the waste and produce
leachate to be collected and treated.
     The emission reduction obtained with flexible membranes depends on the
particular combination of organics and synthetic membrane material.  In a
series of laboratory experiments designed to measure the diffusion of
various organic compounds through polyvinyl chloride (PVC) membranes,
Thibodeaux et al.34 found that  the effectiveness of the membrane as a vapor
barrier varied with the chemical being tested.  The three compounds tested
and their measured fluxes were:
          Methanol--0.58 g/m2/h
     •    Methylcyclohexane--3.43 g/m^/h
     •    1,2-Dichloropropane--13.6 g/m^/h.
                                   4-16

-------
These fluxes are averages of values  reported  in the  reference.   Emission
reductions calculated  using analytical emission models showed  similar vari-
ability when used to calculate emission reductions achieved by 30-mil HOPE.
Calculated reductions  ranged from 0  to 99 percent for three of the model
waste compositions described in Appendix C, Section  C.2.2.35   These
calculations were made using Pick's-  law of diffusion and the diffusion and
permeability characteristics of the  waste constituents and liner.  The wide
range of emission reductions is attributable  to the  wide range of specific
compound permeabilities and the fact that permeation occurs at a fixed rate
for specific compounds.   If a compound is present under the membrane in  low
concentration, then the apparent emission reduction, expressed as a per-
centage, could be low, e.g., 0 percent.  If the compound is present at high
concentration under the membrane, then the apparent  emission reduction will
be high, e.g., 99 percent.  This large range  of variations is  likely to be
typical for wastes put into landfills and wastepiles.
     Reductions in organic emissions depend in part  on the frequency with
which the cover is removed from the  waste to  permit  normal operations such
as waste addition or removal.  The state of repair of the cover  and the
effectiveness of the anchoring system (at the edge of the membrane where it
is secured to the substrate) are other factors that  could influence the
organic emission reduction achieved.  Synthetic membranes can  also be
adversely affected by  sunlight and chemical attack.
     For the purpose of estimating nationwide emissions, emission
reductions of 0, 49.3, and 99.7 percent, for  organic-containing  solid, two-
phase aqueous/organic  (after fixation),  and aqueous  sludge/slurry (after
fixation)  model wastes, respectively, are used for 30-mil HOPE covers.
These values were selected based on  the above described calculations for
three model waste compositions.  Using the same calculation methodology,
emission reductions of 0,  84.8, and  99.9 percent were selected for the same
model  wastes with a 100-mil HOPE cover.   The  emission reduction  calculated
for the two-phase aqueous organic model  waste is also used for organic
liquids and organic sludge/slurry (after fixation) model wastes.  The
calculated emission reduction for the aqueous sludge/slurry is used for
dilute aqueous waste (after fixation).
                                   4-17

-------
4.2  ADD-ON EMISSION CONTROL DEVICES
     Add-on emission control devices are those devices used to separate
contaminants (in this case, organics) from a conveying-waste gas stream.
The add-on devices discussed in this section are carbon adsorbers,
condensers, and several types of combustors.
4.2.1  Generic Control Devices
     Several types of add-on controls may be applied to waste gas streams
and be designed to achieve similar levels of emission reductions.  Given
that equivalent levels of emission reduction may be achieved by the
devices, the choice among the devices for any particular site will depend
on site-specific factors that would tend to make one device less expensive
than the others, or preferable for some other reason.  Subsequent sections
of this chapter will discuss the factors that will influence the choices of
device for particular sites.
     In analyzing the nationwide emission reductions achievable by
applications of the control devices and their associated costs, it has been
necessary to account for the fact that not all control devices can be
applied to all  sources with equal cost and equal effectiveness.  Incompat-
ibility may be a problem for certain waste compositions and certain
controls, for example.  As a result of this fact, the concept of generic
control devices has been developed.  A generic control device is the
combination of specific control devices assumed to be applied to an emis-
sion source category for the purpose of estimating nationwide emissions and
control costs.   For example, a covered tank may be controlled by venting to
any of several  control devices or by equipping it with an internal floating
roof.   Nationwide,  both may be done, so the generic control for fixed- roof
tanks  should be a combination of both.  For purposes of estimating
nationwide emissions and costs, generic control devices must be assumed to
be one of the specific control devices, or some combination of one or more
of those specific control devices listed.
     The add-on control device bases (generic control device definitions)
for estimating  nationwide emission and control costs (control strategies
are described in Chapter 5.0, emission impacts in Chapter 6.0, and capital
                                   4-18

-------
and annualized costs in Chapter 7.0) are presented in Section 4.9 of this
chapter.
4.2.2  Carbon Adsorption
     Gas-phase carbon adsorption is widely used to control organic emis-
sions to the atmosphere.  Carbon's activated surface sites allow organics
to adhere to carbon particles as they pass through a carbon bed.  When the
capacity of the carbon to adsorb organic compounds is exhausted, the spent
carbon is replaced or regenerated.
     The two basic configurations for gas-phase carbon adsorption systems
are regenerative, referred to herein as fixed-bed carbon adsorption, and
nonregenerative, which is referred to as carbon canisters.  In fixed-bed
systems, two carbon adsorption units often are operated in parallel so that
one may operate while the other is being regenerated.  Regeneration of
spent carbon normally is accomplished by steam stripping.  In carbon
canister systems, the spent carbon is replaced with fresh carbon and the
spent carbon is either disposed of or regenerated for eventual reuse.
Detailed descriptions of equipment and operating principles are presented
in several EPA reference documents.36,37,38  /\ simple carbon canister unit,
a 0.21-m3 (55-gal) drum filled with activated carbon and equipped with an
inlet and an outlet port,39 is illustrated in Figure 4-3.
     Activated carbon selectively adsorbs organic vapors from gases even in
the presence of water.41  However, if the gas stream relative humidity
exceeds 50 percent, then the capacity of the carbon for organics may be
diminished.   Also, if the carbon is saturated with water when organic
compound adsorption is begun,  the effectiveness of the carbon is reduced.
Solvents having low molecular weights (particularly Resource Conservation
and Recovery Act  [RCRA] waste code F005 solvents) are not readily recovered
from air streams by activated carbon.42  The presence of multiple organic
components in the gas stream (probably typical for TSDF) may result in
preferential  adsorption of certain components, reducing the apparent
effectiveness of the carbon bed for those components not preferentially
adsorbed.  If the organic constituents likely to be present in the gas
stream are known in advance,  the system design can be adjusted to
                                   4-19

-------
                                          Activated carbon







                                          Support material
Figure 4-3. Nonregenerative carbon adsorption (carbon canister) unit.40
                              4-20

-------
accommodate this situation.  If not, the problem will result in more
frequent regeneration or replacement of the carbon, thus increasing
operating costs.
     For safety reasons, gas streams with organic concentrations greater
than 25 percent of the  lower explosive limit would be precluded from
treatment although dilution air could be added to reduce the concentration.
Additionally, system design should consider the heat released by adsorp-
tion, which could raise the temperature of the carbon bed enough to cause
spontaneous combustion.43  if the waste gas stream contains particulates,
they will tend to plug  the voids in the carbon bed, rendering it
ineffective.  This can  be avoided by filtering the gas feed stream for
particulates.
     Wei 1-designed and  wel1-operated state-of-the-art gas-phase carbon
adsorption systems can  reliably remove 95 percent of many types of organics
contained in hazardous  wastes from contaminated gas streams and are capable
of achieving control efficiencies exceeding 99 percent.44  The adsorption
capacity and thus the efficiency of activated carbon are affected by;45
     •    Type of organic compounds and inlet mass loading
     •    Moisture content of the inlet gas
     •    Temperature of the inlet gas
     •    Carbon type,  amount,  and condition.
     The presence of moisture in the inlet gas stream can decrease the
organic adsorption capacity of the activated carbon.  As a result, waste
gas streams that contain entrained liquids or that have a relative humidity
approaching 100 percent will need to have some of the moisture removed
before the gas stream passes through the carbon bed.  Entrained liquids can
be removed with any one of several types of demisters, and relative
humidity can be reduced by cooling and then reheating the gas stream to
condense out the moisture or by raising the gas stream temperature if the
temperature of the inlet stream is sufficiently low that the increase in
temperature does not affect adsorption efficiency.46  Condensate can be
treated with regeneration steam condensate, as discussed later in this
section.
                                   4-21

-------
     Background Information developed by EPA to support standards for
petroleum refinery wastewater systems indicate that carbon adsorption can
reliably achieve emission reductions of 95 percent.4?  Because of the
similarity of the sources and wastes at TSDF, equivalent emission reduc-
tions would be expected for TSDF sources.  Consequently, in estimating
nationwide emissions, control efficiencies of 95 percent are used for both
fixed-bed carbon adsorbers and carbon canisters.
     A field evaluation of carbon canisters designed to control breathing
and working losses from neutral izer tanks in the wastewater treatment (WWT)
system at a specialty chemicals manufacturing plant is summarized in Appen-
dix F (Section F.2.2.1.2).  In that system, the drums were achieving a high
degree of removal (100 percent) for specific components (i.e., 1,2-
dichloroethane, benzene, toluene, chlorobenzene, and chloroform)  and a
relatively high degree of removal for specific compound groups (except
halogens).
     There are cross-media and secondary environmental impacts associated
with the use of gas-phase carbon adsorption.  Organics that are not
adsorbed from the gas stream will be emitted and may require secondary
control  such as condensation or incineration.  With proper design, i.e.,
knowing what compounds may be present in the gas stream and designing for
them,  this impact should be minimized.  Transfer,  storage, and handling
operation (TSHO)  fugitive emission sources include the handling of spent
carbon if it is regenerated offsite, transport operations to and from
onsite regeneration facilities, and routine maintenance operations such as
column cleaning or equipment repair.  Organic emissions may occur during
off-site regeneration without proper air pollution controls.
     If a steam desorption cycle is used and the recoverable organics are
soluble in water,  then some form of water treatment or separation process
is required for the wastewater.  For example, steam stripping with conden-
sation of the overhead product might be used, which will have associated
process  (e.g.,  the condenser and product receiver vents) and TSHO fugitive
emission sources.   If the organics are insoluble in water, they can be
separated from the condensate by decantation.  Land disposal of spent
carbon constitutes a cross-media impact for this form of emission control.
                                   4-22

-------
     Primary emissions from energy production for carbon regeneration are
those associated with thermal regeneration.  Such emissions include furnace
or afterburner off-gases, which will usually contain the fully oxidized
products of the oxidized adsorbates  (e.g., HC1 from a chlorinated solvent
or SC>2 from a sulfur-containing material).  Also, the regeneration process
will produce an ash that may require further treatment.  If a flue gas
scrubber on the regeneration furnace is necessary, it of course will pro-
duce waste streams that may also need treatment.
4.2.3  Combustion
     4.2.3.1  Thermal Incineration of Organic Vapors.  Vent gases from
fixed-roof storage or treatment tanks and air-supported structures covering
impoundments may be candidates for thermal incineration to destroy the
organic vapors.  Thermal incineration is widely used for organic vapor
destruction in the chemical manufacturing and metal foundry industries.
The extent to which thermal incineration is currently used for vapor
destruction in TSDF is unknown.
     Thermal incineration of organic vapors is achieved in any device that
uses a flame (temperature) combined with a chamber (time and turbulence) to
convert combustible material to carbon dioxide and water.  An incinerator
usually consists of a refractory-1ined chamber that is equipped with one or
more sets of burners.  The organic-laden gas stream is ducted from the
emission sources to the burner zone.  A flame is maintained in the burner
zone by combustion of auxiliary fuel and air.  The organic vapors are
heated above their ignition temperature in the burner zone, then expand
into one or more combustion chambers maintained at a constant temperature.
The combustion products then are exhausted to the atmosphere.  Heat
recovery (e.g., a shel1-and-tube heat exchanger for preheating inlet gas)
may be used to minimize fuel consumption.  Figure 4-4 illustrates a thermal
incinerator.  Detailed descriptions of equipment and operating principles
can be found in several  EPA reference documents.48-51
     Incineration is recognized as the most universally applicable of
available organic emission control methods because it can be used to
destroy essentially all  types of organic compounds from a variety of
sources.52  One possible limitation of an incineration system might be the
                                   4-23

-------
                         Vent stream
Auxiliary fuel
-p*
ro
                            Incinerator
                                                                                          Exhaust stack
                                                                               Scrubber
                                                                                           Scrubber effluent
                         Figure 4-4. Schematic diagram of thermal incinerator system.

-------
necessity of a relatively constant inlet flow rate to sustain stable flame
conditions.53  por a vapor stream with large flow fluctuations, a flare
(discussed in Subsection 4.2.3.3) might be more appropriate than a thermal
incinerator for organic emission control.  Safety may also be a considera-
tion in applying thermal incineration.  In the past, storage tank vents at
hazardous waste incineration facilities were often integrated with the
incinerator feed system.  However, this practice has been discontinued
because of several accidents.54  An additional potential limitation to the
use of thermal incineration would be the presence of sulfur- or halogen-
containing compounds in the vapor stream that would cause corrosion inside
the incinerator.55  in these cases, special construction materials and
additional control equipment may be required to prevent release of corro-
sive combustion products such as hydrochloric acid (HC1) and sulfur
dioxide.
     Destruction efficiency is a function of the following factors:
     •    Inlet waste stream characteristics
     •    Combustion zone temperature
     •    Residence time
     •    Degree of mixing in the combustion chamber.
     Thermal  incineration has been demonstrated to achieve organic
destruction efficiencies of at least 98 percent at a temperature of 870 °C
and a residence time of 0.75 seconds for most compounds.  For halogenated
streams, 98-percent efficiency is predicted for incinerators operated at
1,100 °C with a 1-second residence time.56-61  Consequently, for the
purpose of estimating nationwide emissions, a destruction efficiency-of 98
percent is used for thermal incinerators.
     Organics combustion in a thermal incinerator can produce secondary
emissions consisting mostly of nitrogen oxides.62  However,  these emissions
should be minimal  in a wel1-designed and -operated incinerator.  Combustion
of halogenated organic compounds may release hal-og-e-n-ated combustion
products to the environment although HC1  emissions are normally controlled
by the use of scrubbers.  If a scrubber is used,  the scrubber effluent will
likely require treatment.
                                   4-25

-------
     4.2.3.2  Catalytic Incineration of Organic Vapors.  Vent gases from
 fixed-roof storage and treatment tank or air-supported structures covering
 impoundments are candidates for catalytic incineration.  Catalytic
 incineration is widely used to destroy organic vapors.  One of the most
 common applications is to control hydrocarbon emissions in automobile
 exhaust systems.  The extent to which catalytic incineration is currently
 used in TSDF is unknown.
     Catalytic oxidation (incineration) is a method of controlling organic'
 emissions.  In the presence of a catalyst, organic vapors are oxidized,
 creating carbon dioxide and water.  Catalytic oxidation uses a metal- or
 metallic-alloy-based catalyst to promote higher rates of organics/oxygen
 reactions at lower temperatures than can be achieved with thermal incinera-
 tion.  Many metal oxides can promote catalytic oxidation of hydrocarbons
 and carbon monoxide if the reactant temperature is sufficiently high.  The
 noble metals, in particular platinum and palladium., are most active and are
 used extensively as catalysts, usually in a packed or fixed-bed arrange-
 ment.  Figure 4-5 illustrates a catalytic incinerator.  References 63, 64,
 65, and 66 contain detailed discussions of equipment and operating prin-
 ciples.
     The use of catalytic converters to meet desired emission control has
 been extended to virtually every conventional combustion system as well as
 chemical  processes with partly oxidized components, amines, alcohols,
 acids,  etc.67  /\n advantage of catalytic oxidation over thermal oxidation
 is that less NOX is formed as a result of the lower operating temperature
 and of operating close to the stoichiometric requirement for oxygen.68
Operating limitations mitigate some of the advantages of catalytic oxida-
tion.   Accumulations of particulate matter,  condensed organics, or
polymerized hydrocarbons on the catalyst can block active sites and reduce
efficiency.   Compounds containing sulfur,  bismuth, phosphorus,  arsenic,
antimony,  mercury,  lead,  zinc, tin,  or halogens can combine chemically with
the catalyst,  rendering it useless.   In addition,  catalytic incinerators
are limited to  a narrower range of inlet fuel concentration and temperature
than are  thermal  incinerators.
                                   4-26

-------
                            Vent stream
                                                .
                                            X
                                                 Catalyst bed
Auxiliary
fuel
                             Catalytic
                            incinerator
Preheater
heat
exchanger
Exhaust stack
                                                                                              ^- Scrubber effluent
                             Figure 4-5.  Schematic diagram of catalytic incinerator system.

-------
     Catalytic incineration is a technology that can be transferred
directly from use on other gas streams to use as an add-on control for TSDF
operations.  One potential application of catalytic incineration is to
control overhead organic emissions from an air-stripping column where
humidity might impair the performance of gas-phase carbon adsorption.
     Catalytic oxidation systems can achieve organic destruction
efficiencies approaching 99 percent.  Test data for catalytic systems used
in other industrial applications for organic emission control indicate that
half of the tested units achieved greater than 90 percent destruction of
organic compounds.  The remaining tested units were capable of achieving 80
or 90 percent organic destruction.69  The destruction efficiency depends on
the inlet temperature, residence time, and temperature profile in the
reactor.  Two recent EPA-sponsored studies investigated the effectiveness
of catalytic incinerators as a means of destroying organics and hazardous/
toxic air pollutants.70,71  Both studies concluded that destruction
efficiencies of 97 to 98 percent were achievable for most of the compounds
tested.  The first study indicated that chlorinated hydrocarbons were not
as effectively destroyed as nonchlorinated compounds but the second showed
chlorinated hydrocarbon destruction efficiencies of 97 to 98 percent.  On
the basis of these test results and considering that catalytic incineration
has limited applicability because of the potential 'for catalyst poisoning,
destruction efficiencies of 98 percent appear reasonable for catalytic
incineration and were used in the estimation of nationwide emissions.
     Cross-media and secondary air impacts associated with the use of
catalytic incinerators are minimal, although spent catalyst must occasion-
ally be disposed of.  In applications where the feed stream oxygen is
substantially depleted because of .oxidation of trace contaminants, the same
catalyst can reduce inlet NOX to N2 and 02-  On occasions where fuel is
added to enhance conversion of trace contaminants, a fuel is used that does
not contribute to harmful emissions (e.g., the fuel may be H2 or CH4).
     4.2.3.3  Flares.  Flares commonly are used to incinerate waste gases
from petroleum refining and petrochemical manufacturing operations.  Flares
are used for the direct combustion of waste gases.  A flare consists of a
burner designed to handle varying rates of fuel while burning smokelessly.
                                   4-28

-------
In general, flares can be classified as elevated or ground  level.  An
elevated flare is mounted on a stack so as to remove it from the process
area for purposes of safety  (heat) and nuisance  (noise).  Auxiliary equip-
ment includes a stack seal  (to prevent air intrusion), a knockout drum  (for
removal of entrained liquids), an ignition system, and the  usual ductwork,
capture device, and fan subsystem.72  Figure 4-6 is a schematic of a steam-
assisted elevated flare system.  Ground-level flares are more versatile
than elevated flares but have a much higher capital cost.   Ground-level
flares generally consist of  several burners and  combust the material in a
refractory chamber or an open pit.74  Several EPA  reference documents^, 76
present detailed information on equipment and operating principles.
     Flares can be used for  streams with fluctuations in organic
concentration, flow rate, and inert content.77   However, for most efficient
operation of a given flare design, there are certain limitations to flare
exit velocity and heat content of the waste gas, as described below.  Waste
streams containing high concentrations of halogens should not be combusted
in flares to prevent flare tip corrosion and secondary emissions of SOX or
HC1.78
     Flare destruction efficiency is determined  by characteristics of the
waste gas such as flammability limits,  autoignition temperature, and den-
sity and by flare operating  conditions including flame temperature, resi-
dence time in the combustion zone, and mixing at the flare  tip.79,80  j^e
EPA sponsored a series of experimental  tests to  measure flare destruction
efficiency.81-82,83  in these tests, flare destruction efficiency was
measured under a wide range  of operating conditions.  Two important
conclusions of the studies were:
     1.   When flares are operated under conditions that are represen-
          tative of industrial practices,  the combustion efficiencies
          of the flare plume are greater than 98 percent.84
     2.   Destruction efficiencies of 98 to 99 percent or greater are
        ---achieved when flares are operated under stable flame condi-
          tions.85
     The EPA has translated these conclusions into specific minimum
requirements for flare designs for controlling equipment leaks in the
                                   4-29

-------
Gas Collection Header
and Transfer Un«
 I       I
                                     St*a» Nozzles
                                      Sas 84iTl«r
                                      Rar* Stack
r
               Knock-out
               On*
                                       Su
         Uater.
                                                                 F1ar« Tip
                                                                            Pilot Burners
                             ^•^
 Ignition D«v1ce


Air Une

 Gas Line
                             Drain
                   Figure 4-6.  Steam-assisted elevated flare system.
                                                                          73
                                           4-30

-------
synthetic organic chemical manufacturing industry.86  Minimum flare exit
velocities have been specified based on the heat content of the waste gas
and on whether the flare  is steam-assisted, nonassisted, or air-assisted.
     There currently are  no known applications of flares to TSDF vent
streams, but if vent streams exist with sufficient heat content, a properly
designed and operated flare should achieve a 98-percent destruction effi-
ciency.  Flares were not  used in the estimates of nationwide emissions, but
they might find application to certain vent streams at TSDF.
     Secondary environmental impacts associated with flare operation con-
sist primarily of air emissions of NOX.  However, measurements at two oper-
ating flares show that emissions of NOX are lower than those for incinera-
tors and considerably lower than those for boilers.87  Low-frequency com-
bustion noise and high-frequency jet noise constitute a nuisance problem
for elevated flares in populated areas.  However, this is relatively easy
to minimize with sound-proof enclosures.
4.2.4  Condensation
     Condensers are currently used in TSDF and waste solvent treatment
facilities to remove organic vapors from conveying gas streams in several
types of equipments.  Vent gases from steam strippers, distillation
processes, and thin-film  evaporators, all of which are used to separate
organic compounds from liquid wastes, are typically sent through condensers
to recover the organics in a more concentrated form.
     Condensation is the  process of reducing a gas or vapor to a liquid by
lowering its -temperature  and/or increasing its pressure.  Condensation
occurs when the partial pressure of a pollutant in a gas stream equals its
vapor pressure as a pure  substance at operating conditions.  Reducing the
temperature of the gas stream is more cost effective than increasing the
pressure and is therefore the usual approach.88  Condensation is relatively
easy if the gas-phase hydrocarbon concentration is high.  When concentra-
tions are low,  condensation at reasonably achieved temperatures can be
difficult.89
     Condensers can be classified as contact and surface condensers.  In a
contact condenser,  the coolant and vapor stream are physically mixed and
                                   4-31

-------
exit the condenser as a single stream.  Because the condensate would pose a
disposal problem or require further processing for recovery of the hydro-
carbon, contact condensers are unlikely to be used at hazardous waste TSDF.
     Surface condensers are usually in the form of shell- and tube-heat
exchangers, such as that shown in the schematic in Figure 4-7.  The vapor
stream flows into a cylindrical shell within which are numerous small tubes
conveying the coolant.  Condensation occurs on the cool  surface of the
tubes; i.e., heat is transferred from the vapor stream to the coolant
through a heat exchange surface.  The rate of heat transfer (q) is a func-
tion of three factors:  total  cooling surface available,  resistance to heat
transfer, and mean temperature differential between condensing vapor and
coolant.  For a single-component vapor stream, this can  be expressed mathe-
matically by: 91
                                q = UA ATm
where
       U = overall heat transfer coefficient
       A = heat transfer surface area
     ATm = mean temperature differential.
In practice, the vapor stream will  contain multicomponents, air,  and at
least one other gas, complicating the design procedure.
     The coolant used in a surface  condenser depends on  the saturation
temperature (dewpoint) of the organics.  If the saturation temperature is
high enough, an air-cooled condenser could be used.  Other coolants include
water at ambient temperature,  chilled water, brines, and freons.   Chilled
water can be used to bring the temperature to as low as  7 °C,  brines down
to -34 °C,  and freons below -34 °C.92  The major pieces  of equipment in a
condensation system include the condenser, refrigeration system,  storage
tanks, and pumps.  References 93, 94, and 95 contain detailed discussions
of equipment and operating principles.
     Condensation processes with significant refrigeration requirements are
being used for the recovery of gasoline vapors at bulk gasoline terminals.
Condensers have been used successfully (usually with additional control
equipment)  in controlling organic emissions from petroleum refining and
                                   4-32

-------
     COOLANT INLET    VAPOR OUTLET
                                                 VAPOR INLET
                        	I	I
               r     A
  COOLANT OUTLET     CONDENSED VOC
Figure 4-7. Schematic diagram of a shell- and tube-heat surface condenser.
                                                                 90
                             4-33

-------
petrochemical manufacturing, dry cleaning, degreasing, and tar dipping.yb
Probably the most significant application for condensers at hazardous waste
TSDF is in the recovery of organics volatilized in organic removal
processes such as steam stripping or thin-film evaporation.
     Condensers are not well suited to treatment of gas streams containing
organics with low boiling points or streams containing large quantities of
inert and/or noncondensible gas such as air,  nitrogen, or methane.  Air
stripper offgas, for example, would be an unsuitable application; however,
steam strippers and distillation-type processes would be appropriate appli-
cations.  Because condensers often operate at temperatures below the freez-
ing point of water, moist vent streams must be dehumidified to prevent the
formation of ice in the condenser.  Particulate matter also should be
removed because it may deposit on the tube surfaces and interfere with gas
flow and heat transfer.  Generally, condensers are not considered for vent
streams containing less than 0.5 percent organics.97
     The organic removal efficiency of a condenser is highly dependent on
the characteristics of the vapor stream entering the condenser and on the
condenser operating parameters, as has been discussed.  Efficiencies of 50
to 95 percent are achievable depending on organic concentrations, specific
compounds present, and condenser temperature.98  Appendix F (Section
F.2.2.3) contains a summary of the results of a field evaluation of a
condenser system used to recover organics that were steam-stripped from
wastewater at a plant producing ethylene dichloride and vinyl  chloride
monomer.  Condenser performance varies with volatility of compounds in
vapors,  temperature,  condenser coolant, and condenser surface area.
Process simulations were performed with the Advanced System for Process
Engineering (ASPEN) for a steam stripper connected to a two-stage condenser
designed to strip at  least 90 percent of the medium volatile compounds
present in one of the model wastes described in Appendix C, Section
C.2.2.99  Using the dilute aqueous-2 model waste, two-stage condenser
efficiencies in the range of 98.4 to 99.99 percent were shown to be achiev-
able for the high, medium,  and low volatile compounds by using a refriger-
ated coolant second stage.   This range of efficiencies was used to estimate
the degree of organic removal from overhead vapor streams coming off steam
stripping,  distillation, and thin-film evaporation processes.   Additional
                                   4-34

-------
discussion of condenser usage with organic removal processes is given in
Section 4.3 of this chapter.
     Cross-media and secondary environmental impacts associated with the
use of a surface condenser are primarily the noncondensible gas effluent.
Because the coolant does not contact the vapor stream or condensate, recov-
ered organics are normally reusable.  If the condensate is not reusable, it
is probably incinerated, therefore causing the adverse environmental
impacts associated with incineration.  Other secondary impacts are associ-
ated with the generation of electricity used for air moving and,  in some
cases, refrigeration.
4.3  ORGANIC REMOVAL PROCESSES AND TECHNOLOGIES
4.3.1  Steam Stripping
     Steam stripping involves the fractional distillation of volatile con-
stituents from a less volatile waste matrix.  Both batch and continuous
steam stripping are commercially proven processes and have been commonly
used to remove organics from aqueous streams such as process wastewater.
Several references discuss steam stripping in detail, including a steam-
stripping manual published by EPA,100 discussions of the theory and design
procedures, 101-104 ancj discussions of applicability to.hazardous
wastes.105-108  jhe basic operating principle of steam stripping  is the
direct contact of steam with the waste,  which results in the transfer of
heat to the waste and the vaporization of the more volatile constituents.
The vapor is condensed and separated (usually decanted) from the  condensed
water vapor.  Ultimate control of organics is accomplished by recycling or
incinerating the condensed organic layer.  A simplified diagram of a steam
stripper is  shown in Figure 4-8.
     Batch steam stripping is used extensively in the laboratory  and in
small  production units where a single unit may have to serve for  many mix-
tures.  Large installations also use batch stills if the material to be
separated contains solids,  tars, or resins that may foul or plug  a contin-
uous unit.   Batch stills are also used to treat materials that are gener-
ated from a  cyclical or batch process.109  Batch processing may offer
advantages  at hazardous waste facilities because the unit can be  operated
                                   4-35

-------
i
(JO
CTl
                      Vent
                      J
          Waste in
                         Storage
                           and
                        feed tank
                                                                                         Vent (optional)
                                                                                        J_
Fe
>v
i
mp
Ffflnpnt -*
ed

*_

• 	

	 '
Trays
or
packing
^





Steam
stripper
«< 	 Steam
                                        (wastewater treatment)
                                                                                              Vent
     Condensed
     liquids
Decanter
-Decanted water
 to feed tank

                                                                                              Vent   Recovered organics
                                                                                                 Storage
                                                                                                  tank
                                                                                                                  Recycle, fuel
                                       Figure 4-8. Schematic diagram of a steam stripping system.

-------
In a manner most  suitable for the particular batch of waste to be stripped.
For example, the  same  unit may be used to remove volatiles from a batch of
wastewater, from  a waste containing solids, or from a high-boiling organic
matrix.  The heat input rate and fraction boiled over can be varied for
each waste type to obtain the recovery or removal desired for the specific
batch of waste.   If the system is cleaned between batches, an aqueous waste
stream may be generated from the rinse water.  This rinse water may be
added to a similar batch to be stripped, accumulated in a separate batch
for treatment, or sent to a WWT unit.  However, wastewater may be generated
from cleaning any organic removal or treatment system and would not be
unique to batch operations.  Continuous steam stripping requires a feed
stream that is a  free-flowing liquid with a negligible solids content.
Solids, including tars and resins, tend to foul the column trays or packing
and heat exchangers.   Consequently, wastes containing solids may require
removal of the solids  prior to processing through a continuous steam
stripper.  Unlike the  batch operation, a continuous steam stripper requires
a relatively consistent feed composition to maintain a consistent removal
efficiency from the waste material.HO  j^g continuous steam stripper may
offer cost advantages  over a batch operation for applications in which
there is little variation in the type of feed and for relatively high
volumes of waste materials.
     The products and  residues from steam stripping include-the condensed
vapors (condensate), noncondensible gases, and^the treated waste or efflu-
ent.  The condensate usually is decanted to remove any separate organic
layer from the aqueous layer with recycle of the aqueous condensate back to
the feed stream.  The  separate organic layer may be recovered and reused as
product or fuel.  If the condensate is a single phase of water containing
dissolved organics, then additional treatment of the condensate may be
necessary for ultimate control  of organics.  Most commercial  processes rely
on the formation of a  separate organic phase and decanting for economical
removal  and recovery of organics.  Noncondensibles in the overhead stream
include gases dissolved in the waste material and very volatile compounds
in low concentrations that are not condensed in the overhead system.  The
noncondensibles leave through the condenser or decanter vent and usually
                                   4-37

-------
are vented to the atmosphere or to an incinerator.  For example, vinyl
chloride and chloroethane in one steam stripping test were found to pass
through the condenser and were vented as noncondensibles to an  inciner-
ator.111  The effluent from the steam stripper should be essentially free
of the most volatile compounds; however, semivolatiles and compounds that
are relatively nonvolatile may still be present in the stripper bottoms or
effluent and may require additional treatment for removal.
     Steam stripping is applicable to most waste types that have a reason-
ably high vapor-phase concentration of organics at elevated temperatures
(as measured by the vapor/liquid equilibrium coefficient).  These waste
types are commonly found in TSDF> however, the range of TSDF waste types is
so broad that it is not possible to say whether they represent  a high or
low percentage of the total national wastes.  Theoretically,  wastes can be
processed in a batch still  if they can be pumped into the unit  and if they
produce a residue that can be removed from the still.  This batch operation
may be applicable for waste streams generated in relatively low quantities.
However, batch stripping of sludges with high solids content is not a tech-
nology that has been demonstrated and evaluated in full-scale units-at
TSDF.  Consequently, no design or performance data are available for batch
stripping of sludges.  Some of the difficulties that would be associated
with batch stripping of sludges include material handling problems, heat
transfer in the unit, long cycle times,  and unknown performance.  All of
these factors would affect the basic design and operation of the unit.
     Preliminary treatment such as solids removal  or pH adjustment are
often used before wastewater is stripped in a continuous unit.  Continuous
steam stripping has been used routinely in the chemical industry to recover
organics for recycle and to pretreat wastewater for organic removal prior
to the conventional WWT process.  Some common applications include recovery
of ethylene dichloride,  ammonia, sulfur, or phenol for recycle  and removal
of phenol,  mercaptans,  vinyl chloride, and other chlorinated compounds from
wastewater.112  Batch steam stripping appears to be more common at
hazardous waste facilities  because it is adaptable to different types of
wastes that may be received in batches.^3  For any given waste type,
pilot-scale evaluations or trials in the full-scale process may be required
to optimize the operating conditions for maximum removal at the lowest
cost.

                                   4-38

-------
     Removal efficiencies on the order of 95 to  100 percent are achievable
for volatile compounds such as benzene, toluene, and one- or two-carbon
chlorinated compounds.H4,115  Batch operations  usually provide a single
equilibrium stage of separation, and the removal efficiency is determined
essentially by the equilibrium coefficient and the fraction of the waste
distilled.  The efficiency of a continuous system is related to the
equilibrium coefficient and the number of equilibrium stages, which is
determined primarily by the number of trays or height of packing.  The
organic removal efficiency also is affected by the steam input rate, column
temperatures, and, in some cases, the pH.  Temperature affects the solubil-
ity and partition coefficient of the volatile compound.  The liquid pH also
may affect the solubility and treatability of specific compounds, such as
phenol.   In principle, the removal efficiency in a multistage system can be
designed  to achieve almost any level.  In practice, removal efficiencies
are determined by practical limits in the column design (such as maximum
column height or pressure drop) and cost.  Consequently, steam stripping is
difficult to characterize in terms of maximum achievable performance with
respect to percent organic emission reduction or organic concentration in
the treated waste.
     Several evaluations of steam stripper performance have been reported
in the literature.  The results of five such evaluations are presented in
Appendix  F.  Wastewater containing methylene chloride, chloroform, and
carbon tetrachloride was treated by steam stripping at a chemical plant
(Section  F.2.3.1.1)..  An inlet concentration of  approximately 6,000 ppm
organics was reduced to less than 0.037 ppm for  an overall removal effi-
ciency of about 99.999 percent.  The effluent from the stripper required no
further treatment and was discharged directly to a river under a National
Pollutant Discharge Elimination System (NPDES) permit.
     The steam stripper at one plant (Section F.2.3.1.2) was used to strip
VO, as defined by purge-and-trap analytical procedures, from industrial
wastewater.  The major component was 1,2-dichloroethane, which was removed
from the wastewater at an efficiency of 99.998 percent.  The removal of
total  organics, which generally included chlorinated compounds with one to
two carbon atoms,  averaged 99.8 percent.
                                   4-39

-------
     The test at a chemical manufacturing plant (Section F.2.3.1.3) evalu-
ated the removal of nitrobenzene and nitrotoluene from wastewater.  These
compounds are less volatile than the compounds in the wastewater at the
first plant.  The removal efficiency for nitrobenzene and 2-nitrotoluene
ranged from 91 to 97 percent with an overall organics removal of 92 per-
cent.
     The steam stripper at a fourth chemicals manufacturing plant  (Section
F.2.3.1.5)  is used to remove relatively volatile compounds (methylene
chloride, chloroform, and carbon tetrachloride) from wastewater.  The
removal of  the major component, methylene chloride, was 99.99 percent.  The
removal of  total organics was 98 percent.
     Four batches of waste were evaluated in a batch steam stripping proc-
ess used to reclaim organic solvents (Section F..2.3.1.4).  The types of
compounds present in the waste included both very volatile compounds and
some considered to be semivolatiles because of their solubility in water.
The removal of the most volatile compounds was on the order of 99 percent
with occasionally lower values for specific compounds (e.g.,  91 percent for
acetone, 87 to 94 percent for 1,1,1-trichloroethane, and 74 percent for
ethyl benzene).  The removal of total organics from the batches ranged from
94 to 99.8  percent.
     At a solvent recycling plant (Section F.2.3.4.1), tests were conducted
on two batches of waste processed through a batch steam distillation unit.
In the first batch, removal of individual compounds ranged from 36 to 92
percent and total organics removal  was 76 percent.  In the second batch,
removal of  individual compounds ranged from 12 to 91 percent and total
organics removal was 91 percent.   In this latter batch, a major portion of
the total organic content of the waste consisted of the most volatile
compound.
     Steam stripping removal efficiencies of 99.99, 94.5, and' 16.5 percent
for high, medium, and low volatile compounds, respectively, were used to
estimate the effects of organic removal from wastes on nationwide emis-
sions.   These removal efficiency estimates were developed using the ASPEN
process simulation computer program for a steam stripper applied to the
dilute aqueous-2 model  waste described in Appendix C, Section C.2.2.H6
                                   4-40

-------
The design goal for the steam stripper was to achieve at  least 90 percent
removal of the medium volatile components of the model waste.
     Emission sources associated with a steam stripping operation include:
(1) tank vents (e.g., feed storage, solids removal, decanters, condensate
receivers),  (2) the overhead condenser vent, and (3)  leaks from transfer
and handling operations (pumps, valves, flanges).  The effluent or bottoms
from the steam stripper also may be emission sources  if they contain any
residual volatiles or an appreciable quantity of semi volat-i les, especially
if the effluent is subsequently "processed through an  aerated system or
placed in an impoundment with a long retention time.
     The major secondary air impact associated with steam stripping is the
generation of emissions in the production of steam.   Emissions from indus-
trial boilers include particulate matter, methane and other organics,  SOX,
CO, and NOX.  The quantity of these pollutants generated  is a function of
the amount of steam produced and the type of fuel used in the boiler (e.g.,
No. 2 fuel oil, No. 6 fuel oil with high or low sulfur content, and natural
gas).  Other secondary emissions such as flyash and scrubber effluent are
also produced from the generation of electricity.
4.3.2  Air Stripping
     Air stripping is a process that uses forced air  to remove volatile
compounds from a less volatile liquid.  The contact between air and liquid
can be accomplished in spray towers, mechanical or diffused-air aeration
systems, and packed towers.H?  The focus of this section is on packed
tower air strippers because the vapor-laden air can be sent to a control
device for ultimate control of organic air emissions, whereas the other
devices rely upon dilution in ambient air to avoid environmental  problems.
In packed towers,  the liquid to be treated is sprayed into the top of a
packed column and flows down the column by gravity.  Air  is injected at the
bottom of the column and rises countercurrent to the  liquid flow.  The air
becomes progressively richer in organics as it rises through the column and
is sent to a control  device to remove or destroy organics in the air
stream.   See Figure 4-9 for a schematic of a typical  air  stripping system
with gas-phase organic emission control.
                                   4-41

-------
                                                             Overhead vapors
                                                                                              Control device
                                                                                                          Vented air
Vent
                                              Feed
               J
  Waste in
-p.
i
-^
ro
                   Storage
                     and
                  feed tank
                  Pump
                                                                                      Control device residue
                                                                                      (e.g., spent carbon)
                                                      Air
                                                      stripper
                                       Effluent
                                                                 Air
                                        Figure 4-9. Schematic diagram of an air stripping system.

-------
     The principle of operation  is the equilibrium differential between the
concentration of the organics  in the waste and the air with which  it  is in
contact.  Consequently, compounds that are very volatile are the most
easily stripped.  The .packing  in the column promotes contact between  the
air and liquid and enhances the mass transfer of organics to the air.  The
residues from air stripping include the organics-laden air that must  be
treated and the water effluent from the air stripper.  This effluent  will
contain very low levels of the most volatile organic compounds; however,
semivolatile compounds that are not easily air stripped may still  be
present and may require some form of additional treatment before final
disposal.  The process does not offer a significant potential for  recovery
and reuse of organics.  Condensers generally are not used to recover  the
stripped organics because of the large energy requirements to cool the
large quantity of noncondensibles (primarily air) and to condense  the
relatively low vapor-phase quantities of organic compounds.  Thermal  and
catalytic incinerators and carbon adsorption units are the most common
control devices used for control of the overhead gas stream from air
strippers.  Fixed-bed carbon adsorption systems offer some potential  for
recovery of organics; however, the decision on type of control (organic
destruction or recovery) is usually based on economics.
     Air stripping has been used primarily on dilute aqueous waste streams
with organic concentrations that range from a few parts per billion to
hundreds of parts per million.  The feed stream should be relatively  free
of solids to avoid fouling in the column; consequently, some form  of  solids
removal may be required for certain aqueous hazardous wastes.  In  addition,
dissolved metals that may be oxidized to an insoluble form should  be
removed.  Equipment may be designed and operated to air-strip organics from
sludges and solids in a batch operation; however, this application has not
been demonstrated extensively and is not a common practice.  The major
industrial application of air stripping has been in the removal of ammonia
from wastewater.H8  jn recent years, the use of air strippers has become a
widely used technology in the removal of^volatile compounds from contami-
nated groundwater. H9,120
                                   4-43

-------
     Packed towers can achieve up to 99.9 percent removal of volatiles from
water.121  The major factors affecting removal efficiency include the
equilibrium between the organics and the vapor phase (usually measured by
Henry's  law.constant for dilute aqueous wastes) and the system's design,
which determines mass transfer rates.  Removal efficiency increases as the
equilibrium coefficient increases; consequently, the extent of removal is
strongly affected by the type of waste and the volatility of the individual
VO constituents.  Mass transfer rates (and removal efficiency) are also a
function of the air:water ratio, height of packing, and type of packing.122
The operating temperature is also an important variable that affects
efficiency because of its direct effect on the vapor/liquid equilibrium.
Higher temperatures result in higher vapor-phase concentrations of VO and
higher removal rates.  Air strippers have operational difficulties in
freezing weather that may require heating the input waste stream, heating
and insulating the column, or housing the operation inside an enclosure.
Air strippers are typically designed to remove key or major constituents.
Compounds more volatile than the design constituent are removed at or above
the design efficiency, and less volatile compounds are removed at a lower
efficiency.  Numerous vendors are available for the design and installation
of air strippers.  As is the case with steam strippers, these vendors
usually  require pilot-scale tests on the actual waste material to design
the column and to guarantee minimum removal efficiencies.
     Emission sources associated with an air stripping operation include
tank vents (storage or feed tanks, preliminary treatment tanks) and equip-
ment used to transfer and handle the waste (pumps, valves, etc).  The air
leaving the stripping column usually is treated by incineration (with
destruction efficiencies of 98 percent or higher) or carbon adsorption
(with removal efficiencies of 95 percent or higher if carbon breakthrough
is monitored).  The choice between incineration and carbon adsorption
depends on the specific conditions at the facility.  For example, high
relative humidity in the air stream leaving the air stripper may adversely
affect the adsorption capacity,of a carbon bed.  This could be avoided by
choosing incineration.  However, if the air stream contains chlorinated
                                   4-44

-------
organics, the incinerated air stream may need to be scrubbed to remove HC1,
leading to higher costs.  In this case, it might be better to choose carbon
adsorption and design to avoid the humidity problem.
     The effluent from the air stripper may be an emission source for
semivolatiles that are not removed efficiently, especially if subsequent
processing includes placement in an evaporation pond or disposal impound-
ment.  Air stripping could be used to reduce organics from wastewater prior
to a wastewater treatment operation.
     An air stripper was evaluated at a Superfund site (Section F.2.3.2.1)
where it is used to remove organics from the leachate collected at the
site.  The evaluation focused on optimizing the removal efficiency for
organic components that represented a relatively wide range in volatility.
During one test, the removal of the most volatile constituents (1,2,3-
trichloropropane and xylene) ranged from 88 to 98 percent.  The removal of
semivolatiles such as aniline, phenol, methylphenol, and -ethylbenzene
ranged from 53 to 70 percent.  The removal of total organics averaged
99 percent.
     Air stripping is not specifically used to evaluate the effects of
organic removal  from wastes on nationwide emissions.  Because most air
stripping applications have been to streams containing less than a few
hundred parts per million organic, its usefulness in reducing TSDF organic
emissions may be limited.  However, there may be some site-specific
situations nationally that would be appropriately treated by air stripping.
     Cross-media and secondary air impacts from air stripping are much less
than those from steam stripping because less energy is required.  The major
utility is electricity for blowing air and pumping the liquid, plus
auxiliary fuel needed for incineration or steam needed for carbon regenera-
tion.  Secondary emissions are generated from coal-fired generators that
produce electricity and include particulate matter, CO, SOX, and NOX.
Cross-media impacts would include small quantities of ash if the vent
stream is incinerated and spent carbon if the vent stream is treated with
carbon adsorption.
4.3.3  Thin-Film Evaporation
     Thin-film evaporators (TFE)  are designed to promote heat transfer by
spreading a thin-layer film of liquid on one side of a metallic surface
                                   4-45

-------
while supplying heat to the other side.123  yne unique feature of this
equipment is the mechanical agitator device, which permits the processing
of high-viscosity liquids and liquids with suspended solids.  However, if
solid particles are large, a coarse filtration operation may be required to
pretreat the waste stream going to the TFE.  The mechanical agitator pro-
motes the transfer of heat to the material by exposing a large surface area
for the evaporation of volatile compounds and agitates the film to maintain
the solids in suspension without fouling the heat transfer area.  Heat "can
be supplied by either steam or hot oil; hot oils are used to heat the mate-
rial to temperatures higher than can be achieved with saturated steam
(>100 °C).  TFE can be operated at atmospheric pressure or under vacuum as
needed based on the characteristics of the material treated.  A TFE is
illustrated in Figure 4-10.
     The two types of mechanically agitated TFE are horizontal and
vertical.  A typical unit consists of a motor-driven rotor with longitudi-
nal blades that rotate concentrically within a heated cylinder.  The rotat-
ing blade has a typical tip speed of 9 to 12 m/s and a clearance of 0.8 to
2.5 mm to the outer shell.  In a vertical design,  feed material enters the
feed nozzle above the heated zone and is transported mechanically by the
rotor and grating down a helical path on the inner heat transfer surface
while the volatile compounds are volatilized and leave the evaporator on
the top.  The vapor-phase products from TFE are condensed in a condenser,
and the bottom residues are collected for disposal.
     TFE have been used widely for many years in a number of applications
such as processing of chemicals, Pharmaceuticals,  plastics, and foods.124
Because of their unique features, their use in chemical and waste material
processing has expanded rapidly.  The flexibility in operating temperature
and pressure add potential to TFE for recovering low-boiling-point organics
from a complex waste matrix.
     Although TFE can be used to remove varying levels of organics from a
waste stream,  when applied to hazardous petroleum refinery sludges, the
most suitable mode of operation is to evaporate the water and volatiles and
leave most  of the hydrocarbons that are less volatile than water.
                                   4-46

-------
                                                                          Vent
Vent
      Storage
       and
     feed tank
                      Pump
                                          Thin-film
                                         evaporator
                                                                 Condenser
                                                                                               Decanted
                                                                                               water
                                                                                                 Recycle
                                                     Bottom stream
                       Figure 4-10.  Schematic diagram of a thin-film evaporator system.

-------
     With this mode of operation, the TFE bottom residue contains only low
concentrations of both volatile and semivolatile organic compounds and thus
has a low potential for air emissions after ultimate disposal.  This mode
of operation was used during a pilot-scale test discussed below.  Waste
forms suitable for TFE treatment include organic liquids, organic sludge/
slurry, two-phase aqueous/organic liquids, and aqueous sludges.  TFE would
not be used as a means of treating dilute aqueous waste because of the high
water content in the waste.
     Although TFE technology is readily available,  as with other organic
removal techniques, a pilot-plant study is usually conducted before full-
scale operation to determine the suitability of the TFE for pretreating a
particular waste stream and to identify optimal operating conditions.  The
EPA recently sponsored a pilot-scale test to assess the performance of a
TFE in removing organics from the different types of petroleum refining
wastes.125  jn that study, 98.4 to 99.99 percent of the volatile and 10 to
75 percent of the semivolatile compounds were removed from the sludge.
These results suggest that a TFE can be used to reduce organics
substantially in refinery sludges that are currently land treated.  No
commercial-scale TFE installations have been identified that process the
types of wastes normally handled by TSDF.  However, two installations of
TFE used to recover organics from waste streams have been documented in the
literature and may have some relevance for TSDF operations. 126  jn one
installation (Section F.2.3.3.3), a hazardous waste recycling plant
operates a TFE under vacuum to separate approximately 95 percent of a feed
stream that consists of waste oils,  a small amount of solids, and
approximately 5 percent organics.  In that operation, toluene is removed
from the oily wastes at less than 85 percent efficiency while both
chloroform and methylene chloride are removed at greater than 99 percent
efficiency.  However,  that installation reportedly has significant organic
air emissions through the vacuum pump vent although no estimate was given
for the magnitude of the emissions.   At another organic solvent reclamation
and recycling plant (Section F.2.3.3.2), a TFE operating at atmospheric
pressure was able to remove approximately 76 percent of the acetone and
30 percent of the xylene from a contaminated acetone waste.  Air samples
from the process vent at that operation indicated that air emissions were
                                   4-48

-------
negligible.  At  a  solvent  recycling plant  (Section  F.2.3.3.1),  a  TFE  showed
removal efficiencies of 45 to 99 percent for  individual  volatile  and
semivolatile compounds and yielded a total  organic  removal  efficiency  of  74
percent.
     Factors likely to affect or limit the  applicability or removal effi-
ciency of  TFE include:
     •     Large  changes in the properties of  the waste being treated,
           which  could cause fouling of the  TFE  unit.
     •     The requirement  for separation of water and condensed organ-
           ics when water is evaporated from the waste stream, which
           adds to  the operating expense of  the  unit.
     For the purpose of estimating nationwide emissions, removal  efficien-
cies used  for TFE  are 99.8, 65.9, and 20.7  percent  for high, medium, and
low volatile compounds, respectively.  These  values are based on  the EPA-
sponsored  pilot  test referenced above.  This  test was judged to be more
representative of  TFE operations at TSDF than the other data cited.
     Emission sources associated with the operation of a TFE include
(1) tank vents (blending or storage tanks and product accumulation tanks);
(2) the overhead condenser vent; and (3) fugitive leaks from pump seals,
valves, etc.  The  condensate recovered from TFE is  a concentrated organic
and may produce  organic air emissions when  stored and transferred.  The
bottom residue from TFE should have a low organic content;  however, some
semi volatile compounds may remain in the waste  and  could be emitted in
subsequent handling or disposal.
     Secondary environmental impacts are those  associated with the genera-
tion of energy for heating the TFE.  Steam or hot oil produced in indus-
trial  boilers is commonly used to provide TFE heating.  Pollutants from the
boiler could include CO,  SOX( NOX,  and particulate matter;  the quantity
emitted depends  on the type of fuel used and the quantity burned.  When
carbon canisters are used to control  organic  air emissions  associated  .vith
TFE, the ultimate disposal  of the spent carbon canister constitutes a
cross-media impact.
4.3.4   Batch Distillation
     Batch distillation is  a commonly used process  for recovery of organics
from wastes.  Its principal use is  for recovery of  valuable organic
                                   4-49

-------
chemicals for recycling or reuse and the re-refining of waste oil as
discussed later in this section.  Examples of its use show that  it can be
applied to wastes and reduce the organic air emission potential  of those
wastes by separating the volatile compounds from the wastes.  Although it
has been applied to aqueous wastes,  it has been more typically applied to
predominantly organic wastes.
     The simplest form of distillation is a batch operation that consists
of a heated vessel (called the pot), a condenser, and one or more receiving
tanks.  This process is identical in principle to batch steam stripping
except that the waste charge is heated indirectly instead of by  direct
steam injection.  The waste material is charged to the pot and heated to
boiling; vapors enriched in organics are removed, condensed, and collected
in receiving tanks.  The distillation is continued to a cutoff point deter-
mined by the concentration of organics in the condensate or the  concentra-
tion of organics remaining in the batch.  A common modification  is to add a
rectifying column and some means of returning a portion of the distillate
as reflux (see Figure 4-11).  Rectification enables the operator to obtain
products from the condensate that have a narrow composition range.  Differ-
ent distillate cuts are made by switching to alternate receivers, at which
time the operating conditions may be changed.  If the distillate is
collected as one product,  the distillation is stopped when the combined
distillate reaches the desired average composition.127  Several  references
are available that discuss the design and operation of batch distillation
units.128-134  jne batch still  is operated at a temperature determined by
the boiling point of the waste, which may increase with the time of
operation.   The distillation can be carried out under pressure or under
vacuum.   The use of vacuum reduces the operating temperature and may
improve product recovery,  especially when decomposition or chemical
reaction occurs at higher temperatures.
     Batch  distillation provides a means for removing organics from a waste
matrix and  recovery of the organics  by condensation for recycle, sale as
product,  or for fuel.   The products  and residues include the condensate
that is  enriched in organics and recovered, noncondensibles that escape
through  the condenser vent,  and the waste residue that remains in the pot.
                                   4-50

-------
     Wastes •
     in
I
en
                                                                     Vapors
                                                            Column
                   Vent
Storage
 tanks
                                        Feed
^
                            Pump
                       Heat
                                                                                             Vent (Optional)
                                                                                      Condenser
                                                                             Reflux
Batch still
                                   Overhead
                                    product
                                                                                -^-Bottoms product
                                                                                    (disposal, recycle, fuel)
                                                                                                           Vent
                                                                                          Collecting
                                                                                          tank         vent
Storage
 tank
              -Recycle, fuel
                                   Figure 411.  Schematic diagram of batch distillation with fractionating column.

-------
The noncondensibles are composed of gases dissolved in the waste and very
volatile organic compounds with relatively low-vapor phase concentrations.
The waste material after distillation may have been concentrated with high-
boiling-point organics or solids that are not removed with the overhead
vapors.  These still bottoms may be a free-flowing liquid, a viscous
slurry, or an organic material that may solidify upon cooling.  If the
waste material contains water, a separate aqueous phase may be generated
with the condensate.  This phase may be returned to the batch'or processed
with additional treatment to remove organics or other contaminants.
     Batch distillation may be used for wastes that have a significant
vapor-phase concentration of organics at the distillation temperature.  If
the waste can be pumped and charged to the still pot and the residue can be
removed from the pot, then the waste is likely to be treatable for organic
removal by this process.  Such wastes include dilute aqueous wastes (the
operation would be similar to batch steam stripping),  aqueous or organic
sludges, or wastes with volatiles in a high-boiling-point organic solvent
or oil.  The batch distillation of sludges has not been demonstrated and
evaluated in full-scale units; consequently,  the processing of sludges in a
batch distillation unit is subject to the same limitations described for
the batch steam stripping of sludges (Section 4.3.1).   Batch distillation
has been used to remove organics from plating wastes,  phenol from aqueous
wastes, recovery and separation of solvents,  and re-refining of waste
oils.135,136  The applicability of batch distillation for a specific waste
type can be evaluated by a simple laboratory distillation to assess
potential organic recovery.  As with other organic removal techniques, the
process may require optimization in a pilot-scale or full-scale system for
different types of wastes to determine operating conditions that provide
the desired distillate composition or percent removal  from the waste.
     Batch stills usually are operated as a single equilibrium stage (i.e.,
with no reflux);  consequently, the organic removal efficiency is primarily
a function of the vapor/liquid equilibrium coefficient of the organics at
distillation temperatures and the fraction of the waste boiled over as
distillate.   The use of a rectifying section yields an overhead product
with a composition that can be controlled by the operator.  The removal
                                   4-52

-------
efficiency for  various waste  types can be  highly  variable because of the
dependence on both  properties of the waste (e.g., organic equilibrium) and
the operating conditions  that are used.  Results  of  tests conducted on a
batch unit at a plant engaged in the reclamation  of  contaminated solvents
and other chemicals  are presented in Appendix  F  (Section F.2.3.4.2).  Those
tests revealed  removal efficiencies on the order  of  99.4 to 99.97 percent
for organics, including compounds such as  methyl  ethyl  ketone, 2,2-dimethyl
oxirane, methanol,  methylene  chloride, isopropanol,  and carbon tetra-
chloride.137  Another test on another unit at  the same  site evaluated the
removal of organics  from  contaminated solvents and showed that removal
efficiencies ranged  from  97 percent (for xylene and  ethyl benzene) to
99.9 percent (for trichloroethane).138  These  results  illustrate that batch
distillation has been used successfully to remove organics from aqueous and
organic wastes  or solvents.   In the estimation of nationwide emissions,
removal efficiencies used for batch distillation  are 99, 18, and 6 percent
for high, medium, and low volatile compounds,  respectively.  The values
were derived from a  material  balance that  was  performed as part of a
process design  and  simulation study using  a model waste stream.139  The
model waste used in  the study is the organic liquid model waste described
in Appendix C,   Section C.2.2.
     Emission sources associated with a batch  distillation operation
include (1) tank vents (blending or storage tanks and product accumulator
tanks), (2) the overhead  condenser/decanter vent, and  (3) leaks from pump
seals,  valves,  etc.  The  organic condensate recovered from the still is
concentrated in organics  and may become an organic emission source when it
is stored,  -transferred, or recycled.  The  waste residue from the still
should have a low volatile organic content; however, semivolatile compounds
may remain in the waste and may be emitted in  subsequent processing or
treatment of the waste.   Losses of organics through the condenser vent can
be controlled when necessary by venting to an  incinerator or other control
device.
     Secondary  environmental impacts are those associated with the genera-
tion of energy  for heating the batch still.  Steam from the burning of fuel
in industrial  boilers is typically used in the industry to provide this
                                   4-53

-------
heating.  Pollutants from the boiler include particulate matter, SOX; CO,
and NOX; the quantity emitted depends on the type of fuel (coal, fuel oil,
or natural gas) and the quantity of fuel burned.
4.3.5  Dewatering
     As used herein, dewatering refers to solid-liquid separation achieved
by filtration or centrifugation.  Such devices normally are characterized
according to the force used to achieve the desired separation.  At TSDF,
solid-liquid separation most often is achieved by filtration rather than
centrifugation.  Filtration is achieved by passing the waste stream through
a filtering medium, often a textile product, using force that may be
applied in any of several ways.  Press-type filters consist of a series of
plates covered with a filtering medium and enclosed in a frame.  Separation
is achieved by filling the void spaces between plates with the input
material and then applying pressure to force the plates together and
generate the desired separation.  Examples of this type of filter include
plate and frame, recessed plate, and pressure leaf filters.140  Filtration
force also may be applied by using atmospheric pressure on one side of the
filter medium while the other side is maintained at greater or lesser than
atmospheric pressure.  Examples of this type of filter include rotary drum
vacuum and rotary drum pressure filters.  In rotary drum vacuum filters,
the driving force is achieved by reducing the pressure inside a rotating
drum to below atmospheric.  The drum is covered with a filter medium that
builds up a cake of solids that contributes to filtration efficiency.  The
rotary drum pressure filter uses the reverse principle of applying greater
than atmospheric pressure to the inside of the rotating drum.  In these
filters, the filtering medium is inside the drum.  Advantages of the vacuum
filter include its adaptability to continuous operation and the ease with
which the filter material can be cleaned and maintained.141  in recent
years,  belt filter presses have become one of the more widely used types
for many applications.  These filters use a combination of gravity and
pressure to apply force across the filter medium.  In belt filter presses,
the input stream is applied to a horizontal moving belt that is covered
with filter material.  Gravity forces cause partial separation of the
liquid from the solids in the stream.  As the belt continues to move, it
approaches a second moving belt, and the two move along together over a
                                   4-54

-------
series of rollers that force the belts closer and closer together, creating
pressure on the material between the two belts.  The belts separate for
solids removal, and the filter medium separates from the underlying
supporting web.  At this point, the filter medium can be washed
continuously.  The ease of continuous washing is one of the primary
advantages of the belt filter press.142  other advantages include the
adaptability to continuous operation and the higher throughputs handled
relative to other types of filters.  Exit streams from a filtering or
dewatering operation include the filtrate, which is mostly free of solids,
and the filter cake, which generally has a sufficiently low moisture
content to be handled using solids handling techniques.  A schematic
diagram of a dewatering system is  illustrated in Figure 4-12.
     Dewatering is applicable to any waste stream that consists of a sludge
or slurry such as petroleum refinery sludges.  When used for this applica-
tion, toxic metals remain in the filter cake, which could continue to be
land treated or may be fixated and landfilled, while the liquid passes
through a separation process where oil (which will contain a large fraction
of the organics) is recovered for  recycle.  Little data have been
identified that can be used to estimate the emission reduction achieved by
dewatering.  However, Chevron Research Company has conducted tests that
indicate as much as 90 percent of  the oil in refinery sludges can be
recovered by dewatering and that oil recovery is improved substantially if
the filtration or centrifugation step is followed by a drying step.143  /\t
an EPA-sponsored test at a Midwest refinery, oil removal  using a belt
filter was found to be 78 percent  for an API separator sludge and
66 percent for a dissolved air flotation (DAF) float.144  jf ^he recovery
efficiencies obtained in the tests can be achieved in full-scale
applications,  an equivalent reduction in emissions from land treatment or
landfill  operations would be expected.  Dewatering devices were not used in
estimating nationwide emissions,  but they might be used in lieu of other
emission  control options in some circumstances.
     Emissions during dewatering would come from pumps,  valves, storage,
and other handling operations and  also from any exposed waste surfaces at
                                   4-55

-------
      Waste
      in
en
cr>
                         Storage
                          tank
                                                                         Control device
                                                                                    Vent

I
ip
Dewatering
device



L





	 ^ Filtrate, tr* IAMA/T

                                                                     Vented dewatering
                                                                         enclosure
                                 Figure 4-12. Dewatering system with enclosed dewatering device.

-------
the dewatering  device.   At  the  test  of  the  belt  filter  press  cited  above,
measured  air  emissions were equal  to 21  percent  of  the  volatiles  in  the API
separator sludge  and  13  percent of the  volatiles  and  22 percent of  the
semivolatiles in  the  DAF float.  When vacuum  filters  are used, emissions
would occur with  the  vacuum pump exhaust.   Control  of emissions could be
achieved  by enclosing the operation  where necessary and evacuating  the
enclosure to  a  capture device such as a  carbon adsorption unit.   Vacuum
pump exhaust  also could  be  controlled with  a  carbon adsorption unit  or
other control device.  If carbon canisters  are used to  control emissions
from dewatering operations,  periodic disposal of  spent  carbon would  be
required  and  would constitute a cross-media impact.   If fixed-bed carbon
adsorption is used, additional  environmental  impacts  would be associated
with disposal or  combustion of  the recovered  organics and condensate, and
with fuel  combustion  used to produce steam  for carbon regeneration.
Specific  impacts  would depend on the combustion process  and type  of  fuel
used.
4.4  HAZARDOUS  WASTE  INCINERATION
     Incineration is  an  engineered process  that uses  thermal oxidation of a
bulk or containerized waste to  produce  a less bulky,  toxic, or noxious
material.  Combustion temperature, residence  time,  and  proper mixing are
crucial in controlling operating conditions.145,146   gf  the several  types
of .waste  incineration systems,  four  are  generally cited  as being  the most
useful and having the greatest  potential for  application to wastes proc-
essed at  TSDF.  They  are liquid injection,  rotary kiln,  fluidized-bed, and
multiple-hearth incinerators.   The type  of  incinerator  selected for  a par-
ticular installation  depends on the  waste type and composition as well as
other factors such as whether the  waste  is  in bulk or containerized.
     Liquid injection incinerators are  versatile  and  can be used  to  dispose
of virtually any combustible liquid  that can  be pumped.  The  liquid  waste
must be converted to gas_p_rior to  combustion.  This change is brought about
in the combustion chamber and is generally expedited  by  increasing the
waste surface area by atomization.147   Liquid injection  incinerators oper-
ate at temperatures between 820 and  1,600 °C.  Gas-phase residence times
range from 0.1 to 2 s.
                                   4-57

-------
     Rotary kilns are versatile units that can be used to dispose of
solids, liquids, slurries, and gaseous combustible wastes.  Rotary kilns
are long,  cylindrical rotating furnaces lined with firebrick or other
refractory material in which solids are combusted by themselves or are
incinerated by combustion of an auxiliary fuel or liquid wastes.  Combus-
tion temperatures range from 870 to 1,600 °C depending on the waste
material characteristics.1^  Solids residence time varies from seconds to
hours, depending on the type of waste.  Unless the kiln is very long (i.e,
provides a larger residence time), some type of secondary burning chamber
usually is required to complete combustion of the solid waste.  The heat
release per unit volume is generally quite low, but the rotary kiln
provides a method of mixing solids with combustion air and can be operated
at temperatures in excess of 1,400 °C that are unavailable in other types
of systems.
     A fluidized-bed incinerator consists of a bed of inert granular mate-
rial fluidized by hot air onto which the waste and auxiliary fuel  is
injected.149  fhe waste in turn combusts and returns energy to the bed
material;  thus, heat release per unit volume is generally higher than for
other types of incinerators.  Fluidized-bed incinerators operate at temper-
atures below the softening point of the bed medium,  usually around 450 to
850 °C.150  7ne residence time is generally around 12 to 14 s for a liquid
waste and longer for solid wastes.  This type of incinerator is suited
particularly to heavy sludge and to certain types of organic/inorganic
mixtures.   The inorganic material will stay in the bed and can be removed
as ash.  Scrubbing of flue gases usually is required to remove fine
particulates,  and subsequent flue gas treatment is required for halogen,
sulfur, and phosphorus compounds.
     The multiple-hearth incinerator was designed for the incineration of
low heat content waste such as sewage sludge.  It generally uses large
amounts of auxiliary fuel  and is large in size.  A multiple-hearth unit
generally  has'Tnree operating zones:  the uppermost hearths where feed is
dried (350 to  550 °C),  the incineration zone (800 to 1,000 °C), and the
cooling zone (200 to 350 °C).151  Exit gases have good potential for heat
recovery,  being around 300 to 600 °C.  Temperatures on each hearth can be
                                   4-58

-------
maintained  using  supplemental  fuel.  Multiple-hearth  units "may be  suitable
for hazardous  sludge  disposal, although  it may be necessary to add  an
afterburner to destroy  unburned hydrocarbons that volatilize on the
uppermost hearths.  Several  incinerator  types are shown  in Figure  4-13.
     The cost  of  operating an  incinerator can be reduced by recovering and
using the heat generated by  the combustion of waste.   Primary heat  recovery
can be employed by using the  incinerator exhaust to preheat the incoming
waste stream.  Secondary heat  recovery,  such as a waste  heat boiler, can
also be used if the production process can make use of the steam generated.
Heat recovery  is  shown  in the  incinerator illustration in Figure 4-13.
     Incineration under proper control and using proper  techniques will
provide total  destruction of  all forms of hazardous organic wastes.152
There are two  basic types of wastes:   (1) combustible wastes,  which will
sustain combustion without auxiliary fuel; and (2) noncombustible wastes,
which usually  contain large  amounts of water or other  inert compounds and
will not sustain  combustion without auxiliary fuel.  Organic liquid and
sludge are most suitable for  incineration because of their heat content.
Aqueous sludge, two-phase aqueous/sludge, and organic-containing solids may
be incinerated with auxiliary  fuel to destruct hazardous organics.  Dilute
aqueous waste would not be suitable for  direct incineration because of its
high water content.
     Of the four  types of incineration,  liquid injection and rotary kiln
have been proven  to destruct hazardous waste and to be commercially avail-
able; fluidized-bed and multiple-hearth  are less frequently used tech-
nologies.  However, fluidized-bed incinerators may have greater potential
because of their  compact design, which results in relatively low capital
cost, and their general applicability to solids,  liquids, gases,  and wastes
containing inorganics.153
     With proper  temperature,  residence  time, and turbulence control, most
incinerators are  able to destruct most hazardous wastes.   An incinerator
system with additional afterburner and flue gas handl trrgTystems can be
operated to achieve 99.99 percent or better destruction or removal effi-
ciency (ORE)154 for most wastes to meet  EPA regulations.   However, the
desired destruction efficiency may not be achievable for some recalcitrant
                                   4-59

-------
                                              PROCESS STEAM
                                                        STACK
     CQMBUSTOR
       LIQUID WAIT*
 STORAGE
   WASTE
CONOmONING

    SUPPORT FUEL
     IF REQUIRED
    ATOUIZINO GAS
                   FUMES
 COMBUSTION
       AIR
Fluidized-Bed Incinerator

                           DISPERSION STACK
                                 \
                            PRECOOLER
SCRU83EO
 GASES
          COMBUSTION CHAMOtH
                                   VENTURI SCRUBBER
                                            WATER TREATMENT

                  Liquid Injection Incinerator
                                        RESIDUE
                     Rotary Kiln Incinerator

         Figure 4-13.  Hazardous waste incinerators.
                            4-60

-------
or halogenated  compounds without  substantial  increases  in  incinerator
operating cost;  therefore, the major  limitations on  incineration would be
the type of waste to be incinerated and the cost of  incineration.  For the
purpose of estimating nationwide  emissions, a destruction  efficiency of
99.99 percent is used for hazardous waste  incineration, which  is the EPA
regulatory requirement for hazardous  waste incineration.
     There are  three major cross-media and secondary environmental impacts
associated with -an  incinerator system:  (1) bottom and  fly ash formation
and handling, (2) scrubber effluent and sludge formation and disposal, and
(3) air emissions from the incinerator exhaust stack.   Hazardous ash or
sludge needs to be  disposed of ultimately  at  RCRA-permitted landfills, or
nonhazardous ash can be incorporated  into  other useful  processes such as
production of cement products; scrubber effluent needs  to  be treated before
discharge; and  the  exhaust gas must meet EPA  emission standards.  Any heavy
metals in the feed  stream would likely show up either in the ash or in the
scrubber sludge.  OSW rules mandate fugitive  particulate controls if the
metal content is significant.

4.5  COKING OF  PETROLEUM REFINERY WASTES
     Delayed coking is a process  used in some petroleum refineries to
recover useful  products from the  heavy ends of the raw  petroleum.  In this
process,  the feed stream enters a fractionator where gas oil,  gasoline,  and
lighter fractions are flashed off and recovered.  The fractionator bottoms
are combined with a recycle stream and heated to reaction temperatures of
480 to 580 °C in. the coker heater.  The vapor-liquid mixture from the
heater then enters  the coke drum,  where the primary coking reaction takes
place.  The coke drum provides the proper  residence time, pressure, and
temperature for coking.  In the coke  drum, the vapor portion of the feed
undergoes further cracking as it passes through the drum, and the liquid
portion undergoes successive cracking and polymerization until it is
converted to vapor  and coke.155  Coking units consist of at least two coke
drums so that one can remain online while coke is removed from the other.
     In removal  of coke from the drum, steam  is first injected into the
drum to remove hydrocarbon vapors, which are cooled to  form a steam-hydro-
carbon mixture.   This is followed by water injection to cool the coke and
                                   4-61

-------
allow removal.  When the coke is cooled sufficiently,  a high-pressure water
jet is used to cut th,e coke into pieces that are then  removed from the coke
drum.
     Coking is an alternative to the land treatment or landfill ing of
petroleum refinery hazardous wastes according to the requirements specified
in the Federal Register.  Coke produced from petroleum refinery hazardous
wastes is exempt from regulation as a hazardous waste, if the waste is
coked at the facility where the waste is generated.  In addition, the coke
cannot exhibit any of the characteristics of a hazardous waste.156
     Refinery sludges can be introduced to a delayed coking operation in
one of two ways.  In one process,  sludge is injected into the coker during
the cool-down period.  In that process, the water content of the sludge
contributes to cooling the coke while organics are cracked into products or
polymerized into coke.  Sludge solids are immobilized  inside the coke.157
This process currently is used at  several refineries.   In the second proc-
ess, sludge is introduced into the coker as a part of  the feed stream by
injecting it into the blowdown system where it is vaporized and  recycled.
In that process, the amount of sludge that can be added to the feed stream
must not exceed some small percentage of the total feed,  and the sludge
must undergo extensive dewatering  prior to entering the coking opera-
tion. 158  Only one refinery is known to use this operation.
     No emission measurement data  were found for coking operations;
however, because the entire operation is enclosed, organic air emissions
are estimated to be quite low.  Some emissions would be expected from
transfer, storage, and handling operations associated  with coking.  Most of
these would be expected to come from the transportation of sludge from the
point of generation to the coking  operation and from storage of sludge at
the coker.  Although no definitive data are available  to permit  estimates
of the emission reduction that would be achieved by processing refinery
sludges through a coker rather than land treating, reductions approaching
100 percent are expected.  Increased emissions from the coking operation as
a result of introducing wastes into the feed stream are estimated to be
negligible.
                                   4-62

-------
     Coking  is  a  possible control option at  refineries  that produce  fuel-
grade coke.  At refineries that produce high-quality, electrode-grade coke,
the quality  degradation caused by sludge injection may  be  unacceptable.
For refineries  that do not have an existing  coking operation, coking would
not be a practical emission control alternative.
     Coking may be a viable alternative for  the disposition of hazardous
refinery sludges  for some facilities.  Because this control option is
available only  at facilities that already have a coking operation and
because organic air emissions from coking are estimated to be low, cross-
media and secondary air impacts of adding waste to the  coking operation are
estimated to be negligible.
4.6  PROCESS MODIFICATIONS AND IMPROVED WORK PRACTICES
4.6.1  Submerged  Loading
     Loading emissions generated during waste transfer  are the primary
source of evaporative emissions from waste containers (e.g., tank trucks
and drums) and  tanks.  It occurs when waste  is transferred into a receiving
container and displaces an equal volume of air saturated or nearly satu-
rated with organics.  The quantity of organic air emissions is a function
of the loading method.
     In splash  loading, the influent pipe dispensing the waste is lowered
only partially  into the container or tank and waste is  injected above the
liquid level in the container or tank.  Significant turbulence and vapor-
liquid contact  occur during splash loading, which results in high levels of
vapor generation  and loss.  Control of loading emissions can be accom-
plished by using  submerged loading,.  During submerged loading, the influent
pipe opening is below the liquid surface level,  which decreases turbulence
and evaporation and eliminates liquid entrainment.  The most recent use of
a waste container or tank is also important in determining loading emis-
sions.   For example, if the most recent use of the container or tank was to
store nonvolati]e_waste or if the container or tank has just been cleaned,
it will  contain vapor-free air.  If the most recent use of the container
was to store volatile waste and it has not been cleaned or vented
                                   4-63

-------
(dedicated service), the air in the container or tank will contain organic
vapors that are expelled during the loading operation along with newly
generated vapors.  By changing from top-splash to submerged loading,
organic air emissions can be reduced up to 65 percent when loading clean
tank trucks and 59 percent when loading a dedicated normal service
truck.159  similar emission control efficiencies should be achievable by
submerged loading of drums and stationary tanks.
     In estimating nationwide emissions-, an emission reduction of 65 per-
cent, as derived above from the procedures in Reference 159,  is used for
submerged loading for all waste types and vessel types.  Selection of the
65-percent emission reduction was made on the assumption that, on average,
tank trucks and containers are cleaned before filling.
     Submerged loading of open area sources,  such as surface impoundments
or open-top tanks, is not considered a control technique unless it is used
in conjunction with covers or enclosures over the source.   If the loading
is changed from above to below the liquid surface in the absence of covers
or enclosures, organics that would have been  emitted during filling are
instead emitted quickly from the open liquid  surface by wind blowing across
the source.  However, if the open-top tank is covered or if the impoundment
is enclosed,  then emissions may be generated  from the displacement of vapor
by liquid.  In this case, submerged loading may reduce emissions as
described above for stationary tanks and tank trucks and would provide
additional control of organic air emissions.   There are no cross-media or
secondary environmental impacts associated with submerged loading.
4.6.2  Subsurface Injection
     Subsurface injection is a land treatment practice in which waste is
injected directly into the soil.  The process could be used to apply wastes
to a land treatment site in lieu of surface application.  In subsurface
injection, as opposed to surface application, there is no pooling of liquid
on the soil  surface and thus potentially less opportunity for the material
to be" emitted to the atmosphere.  However, in a field study to evaluate the
relative air  emissions from land treatment plots using surface application
and subsurface injection, no difference in emissions was evident.
                                   4-64

-------
Therefore, subsurface  injection  is not currently being considered for  air
emission control.
4.6.3  Daily Earth Covers
     Daily earth covers, as the  name implies, consist of the daily appli-
cation of an earthen barrier over an emission source.  The technique often
is used to control emissions from landfills.  During normal operation  of a
landfill, waste  is deposited at  specified  locations and then is spread and
compacted with a bulldozer.  If  left exposed, the waste will continuously
emit organics until the supply is exhausted.  To limit these emissions, a
layer of compacted earth usually 15 to 20  cm thick can be installed on top
of the waste.  Ideally, earth covers should be placed as soon as the waste
has been spread  and compacted.
     Typically,  earth  covers are constructed of locally available soil and
most probably soil that had been placed on a spoils pile during construc-
tion of the landfill.  The permeability of the cover is determined by the
size of the pores, the total porosity, and the extent to which the pore
spaces are blocked by  water.  A  dry, coarse-grained soil  may be very inef-
ficient as a vapor barrier, while a wel1-compacted, saturated,  fine-grained
soil may completely block organic air emissions.
     Emission reductions achieved by an earth cover depend on the type of
soil used, its moisture content, the degree of compaction achieved,  and the
thickness of the soil  layer.  Using an analytical  emission model  to esti-
mate emissions through a 20-cm soil layer, calculated emissions changed by
more than two orders of magnitude as soil moisture content was varied from
near zero to saturation.  In practice, emission reductions achieved by
earth covers would vary depending on the impact of local  climate conditions
on soil  moisture content.
4.6.4  Mechanical Mixers
     Mechanical mixers are devices that can be used in the waste fixation
process  for mixing liquid wastes with fixatives such as portland cement,
cement kiln dust, or lime flue dust.  They can serve to replace traditional
open pit fixation operations that currently are used at many TSDF.  One
type of  mixer in current use is the pugmill.161  In general,  a pugmill is
                                   4-65

-------
shaped like a trough roughly 0.9 m wide and 3 m long (the size varies) and
contains two counter-rotating shafts running the length of the unit.  The
shafts are fitted with paddles or cogs designed to agitate and mix the
materials in the device while transporting them from one end of the trough
(where the materials are introduced) to the other end (where the fixed
waste exits).  These mixers process wastes continuously, the throughput
depending on the size and design of the mixer.  Because they can be easily
fitted with covers, emissions from these devices can be captured and con-
trolled more easily than emissions from open fixation pits.
     Potential emission sources associated with mechanical mixer waste
fixation are the storage tank that receives incoming liquid wastes and
sludges, the mechanical mixer, and the processed waste-receiving bin (a
receptacle or area beneath the mixer for temporarily holding fixed waste as
it leaves the mixer and from which the waste is transported to a wastepile
or landfill).  At mixer installations where fixed waste is deposited on a
conveyor belt instead of into a bin, the conveyor belt is an additional
potential emission source.  No data were found that quantify the emission
reduction achieved by replacing open pit fixation with enclosed mechanical
mixers.  However, capture of emissions from mechanical  mixers should
approach 100 percent because of the relatively small size of the mixers and
the ease with which they can be sealed or enclosed (inspection and access
hatches must be kept closed when the mixer is in use).   If 100-percent
capture is assumed, overall control efficiency is a function only of the
efficiency of the control device added on to the vent,  which, for activated
carbon, would be 95 percent or greater (see Section 4.2.2).  In estimating
nationwide emissions, a control efficiency of 95 percent is used for
mechanical mixers vented to a carbon adsorption unit.
     If carbon canisters are used, th-en contaminated carbon disposal must
be dealt with periodically.  Fixed-bed carbon adsorber systems require
steam generation and produce a contaminated condensate that can be fed into
the waste fixation process for final disposal.  If fixed-bed systems are
used,  secondary environmental impacts would be produced from fuel combus-
tion  required to produce steam for regenerating the activated carbon.
                                   4-66

-------
These impacts would consist of combustion products  (e.g., NOX, SOX. CO) and
would vary depending on the type of fuel used.
4.6.5  Housekeeping in Drum Storage Areas
     Drum storage  is the temporary holding of  liquid, semisolid, or solid
wastes until treatment and/or disposal can be  undertaken.  Drums can be
stored on concrete pads that have a perimeter  curb  and gutter for secondary
containment.  Secondary containment is required at  any drum storage area
(40 CFR -264.175),  and spilled or leaked waste  and accumulated precipitation
must be removed  from the sump or collection area in as timely a manner as
necessary to prevent overflow of the collection system.
     Typically,  drums are sealed and in good condition during storage;
therefore, the potential for breathing emissions is assumed to be negligi-
ble.  However, drums may rupture and leak hazardous wastes during storage
or transfer.  Management and technical practices not only cause spillage,
they also determine what fraction is available for  volatilization.  Because
RCRA requires that container storage areas be  inspected weekly for
container leaks  and deterioration (40 CFR, 264.174)., a 50-percent loss of
the volatiles to the atmosphere from the spilled waste was selected for
emission estimating purposes; the remaining 50 percent are recovered as a
result of implementing RCRA spill response actions.
     Two control options are considered appropriate for reducing emissions
from drum storage:  one option is to vent the  existing enclosed drum
storage areas through a fixed-bed carbon adsorption system (see Section
4.2.2 for more detail).  The other control option is to use an open
secondary containment area, conduct daily inspections and maintenance, and
have a policy to clean up spills within 24 hours of discovery.  This option
is an adaptation of the proposed 1986 RCRA tank regulations (40 CFR
264.196).  Leak  detection of stored drums is accomplished by visual
inspection on a  daily basis.  In case of a spill or leak, sorbent material
is used to clean up the spillage within 24 h,  or the spillage is collected
in a collection  sump for transfer to a drum for treatment or disposal
within 24 h.  This policy of daily inspection with  cleanup within 24 h
decreases the fraction of volatiles lost to-the atmosphere.  The magnitude
                                   4-67

-------
of the reduction will depend on operating practices, waste type, and
volatiles concentration; however, no data are available that quantify the
emission reduction achieved by this method.  Because it is estimated that
these housekeeping practices are already used at most TSDF, no credit for
an emission reduction was included in the estimates of nationwide
emissions.
4.7  CONTROL OF EQUIPMENT LEAK EMISSIONS FROM WASTE TRANSFER
     Waste transfer operations often involve pumping waste through pipe-
lines into a variety of waste management process units.  This pumping
creates the potential for equipment leak (fugitive) emissions from pump
seals, valves,  pressure-relief valves,  sampling connections, and open-ended
lines and flanges.  Leaks from these types of equipment are generally
random occurrences that are independent of temperature, pressure, and other
process variables.  However, these leaks do show a correlation with the
vapor pressure of the material in the line.  For example,  monthly inspec-
tion data from the synthetic organic chemicals manufacturing industries
(SOCMI) show that 8.8 percent of the seals on pumps handling light liquids
have leaks and only 2.1 percent of the seals on pumps handling heavy
liquids have leaks.162  Light liquids are defined as those containing at
least 20 percent by weight of organic compounds having a vapor pressure
greater than 10 mm Hg.
     An effective method for controlling fugitive emissions is to implement
a routine leak detection and repair program (i.e., periodic inspection and
maintenance).   Leaks can be detected by individual component surveys, which
may be carried out independently or may be a part of activities such as
area (walkthrough) surveys,  fixed-point monitoring, and visual inspection.
Leaks can be repaired by adjusting the tightness of parts in pumps, valves,
pressure-relief valves, closed-loop sampling, and capping or plugging
open-ended lines or by replacing faulty devices.  The use of portable
organic vapor detection instruments during individual component surveys is
considered to be the best method for identifying leaks of organics from
valves and pump seals;163 use Of suc]n instruments constitutes the only type
of leak detection method for which a control efficiency has been quanti-
fied.
                                   4-68

-------
     The control  efficiency  of  individual  component  surveys  depends  on:
 (1) action  level  or  leak  definition,  (2) monitoring  interval  or  frequency,
 (3) achievable emission reduction  of  maintenance,  and  (4)  interval between
 detection and repair of the  leak.   Background  information  developed  by  EPA
 to support  standards for  SOCMI  fugitive emissions  indicates  that  a monthly
 inspection  and repair program of systems handling  light  liquid reduces
 fugitive emissions from pumps by 61 percent  and  from valves  by 46 per-
 cent. 164  The study also  shows  that closed-loop  sampling and  capping open-
 ended lines provide 100 percent control of these emission  sources.
 Considering the similarity between  SOCMI sources and TSDF  sources, these
 practices are expected to give  equivalent  reductions at TSDF.  Nationwide
 emission reductions were  estimated  based on  the  above  emission reductions
 along with  the estimated  relative  numbers  of pumps,  valves/  sampling con-
 nections, and open-ended  lines  in  the waste management system.  Combined,
 these control techniques  provide weighted  control  efficiencies of 70 per-
 cent for systems  handling light liquids and  78 percent for systems handling
 heavy liquids.
 4.8  CROSS-MEDIA  AND SECONDARY  ENVIRONMENTAL IMPACTS
     When emission controls  described in this chapter  are  applied to an
 emission source,   the controls themselves,  in most  cases, generate cross-
 media or secondary environmental impacts.  Cross-media impacts result from
 any liquid waste, sludges, solid'waste, or air emissions generated by the
 control  device,  and secondary environmental  impacts  result primarily from
 the generation of electricity or steam required by the controls.   The
 preceding discussions of  individual emission controls  include an identifi-
 cation of the types of cross-media  and secondary environmental impacts
 associated with each control.  This section presents information that
 allows" quantification of these  impacts.
     Cross-media  and secondary environmental impacts can be  in the form of
 air emissions, such as SOX,   NOX, CO, and particulates;  aqueous or organic
 liquid waste streams; or sludges or solids.  Air emissions may be generated
by any of the following:
                                   4-69

-------
     •    Burning of fuel oil or natural gas to generate steam for onsite
          carbon adsorber regeneration
     •    Incinerating recovered organics from separation processes or
          process vents
     •    Incinerating auxiliary (supplemental) fuel in an incineration
          process
     •    Generation of electricity.
Aqueous or organic liquid waste streams may consist of:
     •    Condensed steam used in activated carbon regeneration
     •    Decanted organics separated from condensed steam in activated
          carbon regeneration
     •    Exit streams (bottoms) from organic removal processes
     •    Effluent from liquid-phase activated carbon adsorption and steam
          stripping.
Sludges or solids may consist of:
     •    Spent activated carbon
     •    Carbon canisters
     •    Hazardous ash generated by incineration
     •    Ash produced by the combustion of coal to generate electricity
     •    Bottom res.idue from thin-film evaporation
     •    Sludge produced by flue gas scrubbers.
(Quantitative data to support this section are currently under develop-
ment.)
4.9  SUMMARY OF CONTROLS SELECTED FOR DEVELOPMENT OF CONTROL STRATEGIES
     This section summarizes the controls selected for further considera-
tion in the formulation of control  strategies for TSDF air emissions.  (A
control strategy is a combination of specific controls applied to specific
TSDF emission sources resulting in an overall reduction in nationwide TSDF
air emissions.)   The development of control strategies and estimates of the
nationwide emission,  health risk, other environmental, and cost impacts of
                                   4-70

-------
strategies  are  discussed  beginning  in  Chapter  5.0.  Not  all of the emission
controls discussed  in  the previous  sections of this chapter are considered
practical means of  controlling emissions  from  all of the emission sources
identified  in Chapter  3.0.  Table 4-2  lists the TSDF emission sources and
the types of emission  controls that are used to formulate control strate-
gies for reducing TSDF emissions.   The control types are listed by TSDF
process emission source.   The organic  removal/hazardous waste incineration
controls operate on the wastes instead of being applied to a waste manage-
ment process.  The  remainder of the list  includes controls that may be
applied directly to a  waste management process.  Table 4-3 shows the emis-
sion reduction efficiencies used in the nationwide impacts analysis for
specific control options  applied to specific emission sources.  This table
is a summary of the emission reduction efficiencies discussed in preceding
sections and represents the values  used in estimating nationwide impacts of
control strategies.
     Earlier in Section 4.2.1, the  concept of generic control  devices was
introduced.  Although  the control types listed in Table 4-3 for the source
category are typically applicable to the  source category, there may be
site-specific conditions  that prevent the use of a particular control type
at a particular facility.  Where there are several choices of control type
with equivalent levels of  performance for a specific source category,
presumably  the facility operator would choose the lowest cost type to
apply.  In  some cases, there are significant differences in the costs of
control for similar performance levels, but factors such as waste incompa-
tibility or source  shape  may prevent use of the least costly control.
     For this reason,   generic control  devices have been defined for certain
source categories in the  process of estimating nationwide emission reduc-
tions and control  costs.   Table 4-4 lists the options of generic control
devices for the source categories as used in the nationwide impacts esti-
mates.
     Selection  of the percentage weightings for each option was made on the
basis of engineering judgment,  taking into account the application limita-
tions of control options,   relative costs,  and limited information about
                                   4-71

-------
TABLE 4-2.  EMISSION CONTROL OPTIONS USED FOR SELECTING TSDF CONTROL STRATEGIES"
Management process source category
Quiescent surface impoundmentc
Storage or treatment
D i sposa 1

Dumpster storage
Quiescent (storage or treatment) tank
Uncovered

Covered
Waste fixation
Pit
Enclosed mechanical mixer
Aerated/agitated surface impoundment
(treatment)
Aerated/agitated uncovered tank
(treatment)
Land treatment


Active landf i 1 1
Closed landfill
Wastepi le

Control options*5
Organ i c-
removal/HW modi
Suppression Add-on incineration
X
X


X

X
X


X

X


X



X
X
X

X
No direct controls; pretreat to
remove organ i cs or incinerate wastes"



X
X

X
X
X


X
No direct controls; pretreat to
remove organics, incinerate, or
coke wastes"




Process
if ication/work
pract i ces










X











                                                                             (continued)

-------
                                                                TABLE 4-2   (continued)
I
^J
OJ
                        Management  process  source  category
Suppression    Add-on
                                                                                        Control  opt ions"

                                                                                              Organ i c-
                                                                                             removal/HW
                                                                                            i ncineration
     Process
mod i f i cat i on/work
    pract i ces
                        Equipment  leaks
                          Pumps, valves,  and
                          pressure-relief valves
Sampling connections
Open-ended 1 i nes
Drum storage with enclosure'
Drum and tank truck loading
Spi 1 Is
Organic compound removal devices
All TSDF process/emission sources^
X

X
X
X
X
X X
                        TSDF = Transfer,  storage,  and disposal  facilities.
                        HW   = Hazardous  waste.

                        aThe development  of control  strategies  is discussed in Chapter  5.0.
                        ''Emission control  options  are of four types;  examples of  each control  type  are  given  in  Table  4-1.
                         Specific control  options  evaluated in  this study are identified  in  Table 4-3.
                        clncludes treatment impoundments that are dredged annually  and  that  are  thus  exempt from land  disposal
                         restriction regulations.
                        "Disposal surface impoundments function via evaporation.  Direct  controls such  as  covers inhibit
                         evaporation.
                        eCoking is a control  option  only at refineries that have  an existing coking operation  and  that meet  the
                         conditions specified in a November 1985 Federal  Register notice.l"b
                        ^Drums and containers are  stored in a building that can be  vented  to a control  device.
                        9When a waste is  pretreated  to remove organics,  is incinerated, or is processed through  a  coker,
                         emissions from all TSDF emission sources are affected except those  that precede the organic removal,
                         incineration, or coking operation.

-------
                               TABLE 4-3.   EMISSION CONTROL  EFFICIENCIES  USED  IN ESTIMATING NATIONWIDE  IMPACTS
                                                            OF CONTROL  STRATEGIES
      Management process                                                                     	Effect_on_organ j c emi sslons^jj	
       source category                        ControI  dev ice                   Waste  form'3    Capture0   Suppress i on   RemovaI     Contno I
 Quiescent surface impoundment0"     Floating synthetic membrane                    All            —          86
   Storage or treatment             Air-supported structure vented                  All            100         96
                                      bo fixed-bed carbon  adsorber®
 Dumpster  storage                   Dumpster cover                                 AlI            —          99
 Quiescent (storage or  treat-
   ment) tank
   Uncovered                         Fixed rooff                                     All            —        86.4-99.2
                                    Fixed roof with internal  floating              All            —        96.6-99.9
                                      rooff
                                    Fixed roof vented  to carbon canister9           AM            100       86.4-99.2       —           96
                                      or existing combustion  device
                                    Fixed roof vented  to fixed-bed                  All            100      99.3-99.96
                                      carbon adsorber9
   Covered                          Internalfloatingroof^                       3,5,7           —          74-82
                                    Fixed-roof tank vented to carbon                AlI            —          —           —           96
                                      can i sterfl  or ex i st!ng combust i on
                                      dev ice
                                    Fixed-roof tank vented to fixed-bed             All            100         96
                                      carbon adsorberS
 Waste fixation
   Pit                               Fixation with enclosed mechanical               All            100         —           —           96
                                      mi xer  w i th contro I dev i ce
   Enclosed mechanical  mixer         Fixation with enclosed mechanical               All            100         —           —           96
                                      mixer  with control device
 Aerated/ag\tated  surface            A i r-supported structure vented to              AI I            100         95           —           —
   impoundment (treatment)             f i xed-bed  carbon  adsorber"
 Aerated/ag i tated  uncovered          F i xed roof vented  to f ixed-bed                  A tI            100         96
   tank (treatment)                    carbon adsorber
 Land treatment                      Waste coking                                    2   0-          —          —           —           100
 Closed landfill                     30-mlI HDPE  cover                                1                          0
                                                                                   2,3            —          99.7
      _    ______   _    	_	                                    4,5,7           --          49.3
See notes at end  of table.                                                                                                      (continued)

-------
                                                                                TABLE 4-3  (continued)
-P*



m
Manaoemen t process
source category
Wastep i 1 e





Equipment leaks
Pumps , va 1 ves , and
pressure- re 11 ef va 1 ves
Samp 1 i ng connect! ons
Open-ended 1 ines
Drum storage with enclosure1
Drum and tank truck loading
Spi 1 Is
Organ i c compound remova 1 dev i ces




All TSDF process/emission
sources '













Contro 1 dev i ce
100-mi 1 HOPE cover


30-mi 1 HDPE cover




Leak detection and repa i r
C 1 osed- 1 oop samp 1 i ng
Caps or p t ugs
Carbon adsorber
Submerged 1 oad 1 ng
Housekeep i ng
Ex i st i ng contro 1 dev i ceJ
Catalytic inci nerator
Condenser
Therma 1 i nc i nerator
Carbon adsorber
Or pan i c remova 1 and HW i nc i nerat i on

Ai r str ippi ng


Steam stripping


Batch disti 1 lation


Thin-film evaporation


Effect on organic emissionsta %
Waste form'* Capture0 Suppression Removal Control
1 0
2,3 — 99.9
4,6,7 — 84.8
1 0
2,3 — 99.7
4,5,7 — 49.3
Light liquid " 70
Heavy 1 i qu i d



95
All — 65
All — E0
95
98
98.4-99.99k
98
— — — — 95

Constituent volatility
High — — 99.0
Medium — — 13.7
Low — — 1.0
High -- -- 99.999
Medium — — 94.5
Low — — 16.5
High — — 99.0
Medium — -- 18.0
Low — — 6.0
High — — 99.8
Medium — — 65.9
Low — — 20.7
                   See  notes at end of table.
                                                                                                                                                      (conti nued)

-------
                                                           TABLE 4-3  (continued)
     Management process                                                                      	Effect on organic emissions." K	
      source category                        ControI dev ice                   Waste form0    Capturec   Suppression   Remove I      ControI
AlI TSDF process/emission          Hazardous waste  incineration                   AlI            --          —          —         99.99
   sources'  (con.)
                                     Leak detecti on and repa i r, cIosed-      L i ght  1iquid       —          70
                                      Ioop samp ling, and caps or plugs        Heavy  liquid       —          78          --          —
HOPE = High-density polyethylene.
TSDF = Transfer, storage, and disposal facilities.
  HW = Hazardous waste.                                                                                                "
  -- = Not applicable.

3A control device may affect emissions in any of four ways.  It may capture (or contain) emissions and pass them to an emission control
 dev i ce;  it may suppress emissions by containing them or reduc i ng the rate at wh i ch they Ieave the source; i t may remove organ i cs from the
 waste stream before the stream enters other waste management processes; or it may controI  emi ss i ons by destroy i ng the organ i cs or remov i ng
 organ Ics from a vent stream.
'-'Waste form identification:
    1  Organ i c-con'ta 1 n't ng so I id
    2  Aqueous sludge/slurry
    3  D!lute aqueous waste
    4  Organ ic Ii qu i d                                                                                          - _
    6  Organlc sIudge/sIurry
    6  Mi see I Ianeous
    7  Two-phase aqueous/organic.
cCapture efficiencies are based on application of the controls to the emission sources  identified in Table 4-2.
^Includes treatment impoundments that are dredged annually and that are thus exempt from I and d i sposaI restr i cti on reguI at1ons.
eDesorbed mater i a I   is returned to i mpoundment.
°Effi c iencies vary by compound volatiI!ty.
^The efficiency of a fixed-roof tank  in suppressing emissions relative to an open tank  varies depending on the characteristics of the waste
 and the operat i ng parameters of the tank.
9*A fixed roof applied to an open tank suppresses emissions by 86.4 to 99.2 percent.  If the fixed roof is vented to a control device,  the
 remaining emissions are reduced by 95 percent to give an overall emission reduction of 99.3 to 99.96 percent.  If the.control device  is a
 f i xed-bed carbon adsorber, desorbed materia I is returned to the tank so that the tota I emi ss i on reducti on is by suppress i on.  If the
 control  device is a carbon canister or existing combustion device, the overall emission reduction consists of 86.4 to 99.2 percent
 suppression and 95 percent control.
nDesorbed materiaI  is returned to impoundment.
'Drums and containers are stored in a building that can be vented to a control device.
JAn  existing control device is defined as one of the add-on controls described in this  chapter to which the source may be vented.
*Ef f!c!enc i es vary  by compound voI at i Ii ty.
'When  a waste is pretreated to remove organics, or is incinerated, emissions from'all TSDF emission sources are affected
 except those that  precede the organ i c remova I, or i nc i nerati on.

-------
                       TABLE 4-4.  GENERIC CONTROL DEVICE DEFINITIONS3
Management process
  source category
                           ControI
                            type
                     Percentage
                     weight!ng
Control options
      Covered
Quiescent (storage
or treatment) tank

     Uncovered        Suppression plus         50         Fixed roof plus internal floating
                                                          roof
                                               25         Fixed roof plus venting to carbon
                                                          canister or fixed-bed carbon
                                                          adsorber

                                               25         Fixed roof plus venting to existing
                                                          controI  dev i ce

                                               50         Internal floating roof

                                               25         Vent t'o carbon canister or
                                                          fixed-bed carbon adsorber

                                               25         Vent to existing control device

                                               70         Replace pit with mechanical mixer
                                                          vented to fixed-bed carbon
                                                          adsorber

                                               30         Vent existing mechanical mixer
                                                          to fixed-bed carbon adsorber

aThis table defines the combinations of control options used to estimate nationwide emission
 reductions and control costs for the  listed source category and control type.  Potentially
 applicable control options are  listed in Table 4-2.

"Percentage weightings show the  percentages of. each control option emission reduction and
 cost used to define overaI I emission  reductions and costs for the particular combination of
 source category and control type.
  Waste f i xati on
                      Suppression plus
                      suppression or
                      suppression plus
                      add-on
Suppression or
add-on
Add-on

-------
current industry practice.  For example,  venting a tank to an existing
control device onsite is generally less expensive than using an internal
floating roof or venting to carbon canisters.  However, not all TSDF have
an existing control device suitable to receive the tank vent stream.
Depending on number of tank turnovers and organic content of the waste, an
internal floating roof will be less expensive than venting to a carbon
canister.  However, some wastes are not compatible with roof seal  materi-
als, so the internal floating roof is not always applicable either.  With a
low number of tank turnovers and low concentrations of low-volatility
organics, carbon canisters can be cost-competitive with internal floating
roofs  (details of estimated emission reductions and control costs  for the
tank model units storing or treating model  wastes are given in Appendix C).
The percentage weightings in Table 4-4 for tank controls have been selected
after considering the above factors.
     In the case of waste fixation, visits  to a number of waste fixation
facilities have shown that fixation is done both in fixation pits  and 'in
mechanical mixers.  No survey data are available, but pits are believed to
outnumber the mixer operations.  The 70:30  ratio of pits to mixers has been
selected as a best estimate of current status.
4.10  REFERENCES
1.    McCoy and Associates.  Air Supported  Structure for Emissions Control.
      The Hazardous Waste Consultant.  4:1-24.  November/December  1986.
2.    Telecon.  Goldman, Leonard, RTI, with Horn, James, Army Corps of
      Engineers, Tulsa,  OK.  November 25, 1986.  Air-supported domes.
3.    Telecon.  Goldman, Leonard, and Chessin, Robert, RTI, with Weiner,
      Ted, Air Structures, Inc.,  Tappan,  NY.   November 22, 1985.  Air-
      supported structures.
4.    University of Arkansas and Louisiana  State University.  In-Situ
      Methods for the Control  of Emissions  from Surface Impoundments and
      Landfills:  Draft  Final  Report.  Prepared for U.S. Environmental
      Protection Agency.  Contract No. CR810856.  June 1985.  p. 10.
5.    Telecon.  Chessin, Robert,  RTI, with  Weiner, Ted, Air Structures Inc.
      November 22,  1985.  Costs,  life, and  installation of air supported
      structures.
                                   4-78

-------
6.    Reference 4, p. 8 through 9.

7.    Trip report.  Research Triangle Institute.  Visit to the Cecos Land-
      fill, Livingston, LA.  Prepared for U.S. Environmental Protection
      Agency.  September 12, 1986.

8.    Trip report.  Research Triangle Institute.  Visit to Chemical Waste
      Management, Inc., Sulphur, LA.  Prepared for U.S. Environmental Pro-
      tection Agency.  September 12, 1986.

9.    Trip report.  Research Triangle Institute.  Visit to Michigan Dispo-
      sal, Dearborn,  MI.  Prepared for U.S. Environmental Protection
      Agency.  May 29, 1987.

10.   Memorandum from Johnson, W. L., EPA, to list of EPA addresses.
      September 24,  1985.  VOC abatement for small storage tanks (draft).

11.   Memorandum from Yang, S., RTI, to Thorneloe, S., EPA/OAQPS.
      September 4, 1987.  Air emissions for internal floating roof tanks.

12.   Memorandum from Coy, D., RTI, to Thorneloe, S., EPA/OAQPS.
      February 10, 1988.  Documentation of efficiency estimates for
      suppression controls  (draft).

13.   Reference 12.

14.   U.S. Environmental Protection Agency.  Compilation of Air Pollutant
      Emission Factors.  Volume 1:  Stationary Point and Area Sources.
      Chapter 4.3.  Storage of Organic Liquids.  Office of Air and Radia-
      tion, Office of Air Quality Planning and Standards.  Research
      Triangle Park,  NC.  September 1985.  p. 4.3-1 through 4.3-25.

15.   Reference 14,  p. 4.3-3.

16.   American Petroleum Institute.  Evaporation Loss from Internal
      Floating-Roof Tanks, 3rd Edition.  Washington, DC.  Bulletin No.
      2519. 1983.

17.   U.S. Environmental Protection Agency.  VOC Emissions from Volatile
      Organic Liquid Storage Tanks.  Background  Information for Proposed
      Standards.  Office of Air Quality Planning and Standards.  Research
      Triangle Park,  NC.  Publication No. EPA-450/3-81-003a.  July 1984.
      p. 4-9.

18.   Reference 14.

19.   Reference 11.

20.   Reference 12.

21.   Reference 11.
                                   4-79

-------
22.   Reference 17,  p. 4-11.

23.   Telecon.  Chessin,  Robert,  RTI, with Marshall, Harvey, CBI Na-Con,
      Inc., Norcross, GA.  June 9,  1987.  Application of external floating
      roofs to tanks, costs of internal and external floating roofs.

24.   Telecon.  Chessin,  Robert,  RTI, with Nasan, Don,  Pittsburgh-Des
      Moines, Inc.  June 9, 1987.  Costs and applications of external
      floating roofs.

25.   Telecon.  Chessin,  Robert,  RTI, with Janacek,  William.  June 9, 1987.
      Costs and application of floating roofs.

26.   U.S. Environmental  Protection Agency.  Floating Cover Systems for
      Waste Lagoons:  Potential Application at the Old Inger Site.  Draft.
      Oil and Hazardous Materials Spills Branch,  Municipal Environmental
      Research Laboratory.   Edison, NJ.  January  1984.   p. 6-16.

27.   Reference 4, p. 10.

28.   Telecon.  Chessin,  Robert,  RTI, with Matheson, Mike, Gundle Linings,
      Houston, TX.  July 14, 1986.   Floating synthetic membrane costs and
      installation.

29.   Reference 28.

30.   Reference 28.

31.   Reference 4, p. 11.

32.   Telecon.  Gitelman, Amy..,  RTI  with Haxo,  Henry, Matrecon, Inc.
      October 5,  1987.  Estimating  emissions from floating synthetic
      membranes.

33.   Reference 12.

34.   Thibodeaux,  L. J.,  G. J.  Thoma, and C. S. Springer.  The
      Effectiveness  of Poly(vinyl-chloride) as a  Vapor Barrier for Use in
      Hazardous Waste Landfills.   In:  Proceedings of the 3rd International
      Symposium on Operating European Hazardous Waste Management
      Facilities,  Odense, Denmark.   September 16-19, 1986.

35.   Memorandum  from Stallings,  R.,  RTI, to Docket.  February 11, 1986.
      Estimating  the permeability of HOPE to VOC.

36.   Joseph, G.,  T., and D. S. Deachler (Northrop Services, Inc.).  APTI
      Course 415  Control  of Gaseous Emissions — Student Manual.  Prepared
      for U.S. Environmental Protection Agency.  Research Triangle Park,
      NC.  Publication No.  EPA-450/2-81-005.  December 1981.  p. 3-20 and
      p.  5-1 through 5-39.
                                   4-80

-------
37.   U.S. Environmental Protection Agency.  Control Techniques for
      Volatile Organic Emissions from Stationary Sources.  (Draft report.)
      3rd ed.  OAQPS/ESED/CPB.  Research Triangle Park, NC.  March 1986.
      p. 3-28 through 3-51.

38.   U.S. Environmental Protection Agency.  Air Pollution Engineering
      Manual.  Research Triangle Park, NC.  Publication No. AP-40.  May
      1973.  p. 189-198.

39.   Vogel,  G. A., and D. F. O'Sullivan (The MITRE Corporation).  Air
      Emission Control Practices at Hazardous Waste Management Facilities.
      Prepared for U.S. EPA/OSW.  Washington, DC.  Report No. MTR-83 W89.
      June 1983.  p. 35.

40.   Reference 36, p. 5-29.

41.   Reference 37, p. 3-36.

42.   Engineering Science.  Technical Assessment of Treatment Alternatives
      for Waste Solvents--Draft Final Report.  Prepared for U.S.
      Environmental Protection Agency, Technology Branch.  Washington, DC.
      April 1984.   p. 4-16.

43.   Reference 37, p. 3-38.

44.   U.S. Environmental Protection Agency.  VOC Emissions from Petroleum
      Refinery Wastewater Systems — Background Information for Proposed
      Standards.  Research Triangle Park, NC.  Publication No. EPA-450/3-
      85-OOla.  February 1985.  p. 4-30.

45.   Reference 44, p. 4-27.

46.   U.S. Environmental Protection Agency.  Organic Chemical
      Manufacturing.  Volume 5:  Adsorption, Condensation, and Absorption
      Devices.  Research Triangle Park, NC.  Publication No.
      EPA-450/3-80-027.  December 1980.  p. II-2.

47.   U.S. Environmental Protection Agency.  Evaluation of Emission
      Controls for Hazardous Waste Treatment, Storage, and Disposal
      Facilities.  Office of Air Quality Planning and Standards.  Research
      Triangle Park, NC.  Publication No. EPA-450/3-84-017.  November 1984
      p. 111-33.

48.   Reference 36, p. 3-19 through 3-26.

49.   Reference 37, p. 3-7 through 3-10.

50.   Reference 38, p. 171-179.
                                   4-81

-------
51.   U.S. Environmental Protection Agency.  Organic Chemical
      Manufacturing.  Volume 4:  Combustion Control Devices.  Research
      Triangle Park, NC.  Publication No. EPA-450/3-80-026.  December 1980,
      354 p.

52.   Reference 44, p. 4-30.

53.   Reference 44, p. 4-32.

54.   Reference 39, p. 8.

55.   Reference 37, p. 3-9, 3-10.

56.   Reference 37, p. 3-10.

57.   Memorandum and attachments from Farmer, J. R., EPA/OAQPS, to address-
      ees.  August 22, 1980.  Thermal incinerators and flares.

58.   Lee, Kun-chieh, James L. Hansen, Dennis C. McCauley  (Union Carbide).
      Predictive Model of Time-Temperature Requirements for Thermal
      Destruction of Dilute Organic Vapors.  South Charleston, WV.

59.   Midwest Research.  Emission Test of Acrylic Acid and Ester
      Manufacturing Plant,  Union Carbide, Taft, LA.  Draft Report.  May
      1979.

60.   Midwest Research.  Emission Test.of Acrylic Acid and Ester
      Manufacturing Plant,  Rohm and Haas, Deer Park, TX.   Draft Final
      Report.  May 1979.

61.   Midwest Research.  Stationary Source Testing of a Maleic Anhydride
      Plant at the Denka Chemical Corporation, Houston, TX.  Final Report.
      EMB Report 78-OCM-4.   March 1978.

62.   Reference 37, p. 3-12, 3-13.

63.   Reference 36, p. 3-27 through 3-30.

64.   Reference 37, p. 3-13 through 3-18.

65.   Reference 38, p. 179-183.

66.   Reference 51.

67.   Ross, R. D.  Industrial  Waste Disposal.  New Y.ork, Van Nostrand
      Reinhold Co.  1968.

68.   Reference 51, p. 1-3.

69.   Reference 44, p. 4-36.
                                   4-82

-------
70,
71.



72.



73,

74.

75.

76.

77.

78.

79.

80.

81.



82.



83.
84.

85.

86.
U.S.  Environmental  Protection Agency.   Parametric  Evaluation  of  VOC/
HAP  Destruction  via Catalytic Incineration.   Project  Summary.
Research  Triangle  Park, NC.  Publication  No.  EPA/600/52-85/041.  July
1985.  4  p.

U.S.  Environmental  Protection Agency.   Destruction of Chlorinated
Hydrocarbons by  Catalytic Oxidation.  Washington,  DC.   Publication
No.  EPA-600/2-86-079.  September  1986.  p. 9.

Vatavuk,  W. M.   Control Costs.   In:   Handbook of Air  Pollution Tech-
nology, Calvert, Seymour, and Harold  M. Englund  (ed.).  New York,
John  Wiley and Sons.   1984.  p. 365.
Reference 44,
Reference 72,
Reference 36,
Reference 37,
Reference 44,
Reference 37,
Reference 44,
Reference 37,
P-
P-
P-
P-
P-
P"
P-
P-
4-22.
365.
3-16 through 3-19.
3-22 through 3-27.
4-26.
3-26.
4-23 and 4-24.
3-22.
U.S. Environmental Protection Agency.   Flare Efficiency Study.
Research Triangle Park, NC.  Publication.No. EPA-600/2-83-052.
Prepared by Engineering Science,  Inc.   July 1983.  133 p.

U.S. Environmental Protection Agency.   Evaluation of the Efficiency
of  Industrial Flares:  Test Results.  Research Triangle Park, NC.
Publication No. EPA-600/2-84-095.  May  1984.  178 p.

U.S. Environmental Protection Agency.   Evaluation of the Efficiency
of  Industrial Flares:  Flare Head Design and Gas Composition.
Research Triangle Park, NC.  Publication No. EPA-600/2-85-106.
September 1985.  129 p.

Reference 81, p. 5.

Reference 83, p. 2-22.

U.S. Environmental Protection Agency.   Code of Federal Regulations.
Title 40, parts 60 and 61.  Equipment Leaks from Synthetic Organic
Chemical Manufacturing Industry;  Natural Gas Processing Plants;
Equipment Leaks of Benzene Flare  Requirements.  Office of the Federal
Register.  Washington, DC.  July  1, 1986.
                                   4-83

-------
87.   Reference 37,  p.  3-27.

88.   Reference 36,  p.  6-3.

89.   Reference 44,  p.  4-36.

90.   Reference 46,  p.  II-2.

91.   Reference 36,  p.  6-10.

92.   Reference 44,  p.  4-38.

93.   Reference 36,  p.  6-1 through 6-19.

94.   Reference 37,  p.  3-63 through 3-74.

95.   Reference 46.

96.   Reference 37,  p.  3-69.

97.   Controlling Emissions with Flare Towers.   Chemical  Week.   132(21) :49.
      May 25, 1983.

98.   Reference 46,  p.  III-l.

99.   Memorandum from Peterson,  Paul,  RTI,  to Thorneloe,  Susan,  EPA/OAQPS.
      January 18, 1988.  Basis for steam stripping of organic removal
      efficiency and cost estimates used in the Source Assessment Model
      (SAM)  analysis (draft).

100.  U.S. EPA/ORD/IERL.   Process Design Manual for Stripping of Organics.
      Cincinnati, OH.  Publication No. EPA-600/2-84-139.   August 1984.

101.  Schweitzer, P. A.  Handbook of Separation Techniques for Chemical
      Engineers.  New York, McGraw-Hill  Book Co.  1979.  p. 1-147 through
      1-178.

102.  Perry,  R. H.  (ed.).  Chemical Engineers'  Handbook.   5th ed.  New
      York,  McGraw-Hill Book Co.  1973.   p. 13-1 through  13-60.

103.  King,  C.  J.  Separation  Processes.   New York, McGraw-Hill  Book Co.
      1971.   809 p.

104.  Treybal,  R. E.  Mass-Transfer Operations.  New York, McGraw-Hill Book
      Co.  1968.  p. 220-406.

105.  Berkowitz, J.  B., et al.  Unit Operations for Treatment of Hazardous
      Industrial Wastes.   Noyes  Data Corporation.  Park Ridge,  NJ.   1978.
      p.  369-405, 849-896.

106.  U.S. EPA/ORD/HWERL.  Preliminary Assessment of Hazardous Waste
      Pretreatment  as an  Air Pollution Control  Technique.  EPA 600/2-86-
      028, NTIS PB46-17209/A6, March 1986.
                                   4-84

-------
107.   Exner, J. H.  Detoxification of Hazardous Waste.  Ann Arbor, MI, Ann
      Arbor Science.  1980.  p. 1-39.

108.   Metcalf and Eddy,  Inc.  Briefing:  Technologies Applicable to
      Hazardous Waste.  Prepared for U.S. Environmental Protection Agency.
      Cincinnati, OH.  May 1985.  Sections 2.9, 2.15, 2.16.

109.   Reference 102, p.  13-50 through 13-55.

110.   Reference 106, p.  45.

111.   Research Triangle Institute.  Field Test and Evaluation of the Steam
      Stripping Process at B. F. Goodrich, LaPorte, Texas.  Prepared for
      U.S. EPA/ORD/HWERL.  Cincinnati,  OH.  Contract No. 68-03-3253, Final.
      December 5, 1986.   91 p.

112.   Reference 106, p.  43.

113.   Allen, C. C.,  et al. (Research Triangle Institute).   Field
      Evaluations of Hazardous Waste Pretreatment as an Air Pollution
      Control Technique.  Prepared for U.S. EPA/ORD/HWERL.  Cincinnati,  OH.
      Contract No. 68-03-3253.  March 31, 1987.  p. 63.

114.   Reference 100.

•115.   Hwang, S. T.,  and P- Fahrenthold.  Treatability of Organic Priority
      Pollutants by Steam Stripping.  In:  AIChE Symposium Series 197,
      Volume 76.  1980.   p. 37-60.

116.   Reference 99.

117.   Reference 108, Section 2.16.

118.   Reference 105, p.  869.

119.   Reference 108, Section 2.16.

120.   U.S. EPA/Control Technology Center.  Air Stripping of Contaminated
      Water Sources-Air Emissions and Controls.  Research  Triangle Park,
      NC.   Publication No. EPA-450/3-87-017.  August 1987.  125 pp.

121.   Reference 108, Section 2.16.

122.   Reference 105, p.  869-880.

123.   Reference 113, 'p.  23.

124.   Luwa Corporation.   Product Literature--Luwa Thin-Film Evaporation
      Technology.  P.O.  Box 16348, Charlotte, NC  28216.
                                   4-85

-------
125.   Harkins,  S.  M.,  et  al.  (Research  Triangle Institute).   Pilot-Scale
      Evaluation of a  Thin-Film Evaporator for Volatile Organic Removal
      from Petroleum Refinery  Wastes.   Prepared for U.S.  EPA/ORD/HWERL.
      Cincinnati,  OH.   EPA Contract  No.  68-02-3253.  June 1987.  p.  2-1.

126.   Reference 113,  p.  65-84.

127.   Reference 102,  p.  13-50  through  13-55.

128.   Reference 101.

129.   Reference 102,  p.  13-50  through  13.55.

130.   Reference 103.

131.   Reference 104.

132.   Reference 105.

133.   Reference 106.

134.   U.S. Environmental  Protection  Agency.   Distillation Operations in
      Synthetic Organic  Chemical  Manufacturing-Background Information for
      Proposed Standards.   Publication  No. EPA-450/3-83/005a.   December
      1983.  395 p.

135.   Reference 107,  p.  3-25.

136.   Reference 108,  Section  2.9.

137.   Allen,  C. C.  (Research  Triangle  Institute).   Hazardous  Waste
      Pretreatment for Emissions  Control:   Field Tests of Fractional
      Distillation at  Plant B.   Prepared for U.S.  EPA/ORD/HWERL.
      Cincinnati,  OH.   EPA Contract  No.  68-02-3992, Work  Assignment  No. 35
      January 1986.

138.   Reference 137.

139.   Memorandum from  Champagne,  Paul,  and Rogers,  Tony,  RTI,  to Thorneloe
      Susan,  EPA/OAQPS.   March 20,  1987.  Detailed  design and  cost
      memorandum for batch distillation of an organic liquid waste.

140.   Reference 102,  p.  19-57  through  19-85.

141.   Reference 102,  p.  19-76.

142.   JACA Corporation.   Preliminary Assessment of  Predictive  Techniques
      for Filtration,  Drying,  Size  Reduction, and Mixing Unit  Operations.
      Prepared for U.S.  Environmental  Protection Agency.   Cincinnati, OH.
      p.  a-13 through  a-14.
                                   4-86

-------
143.  Northeim, C.  (Research Triangle Institute).  Summary of EPA/Chevron/
      API/RTI meeting on pretreatment of land treatable wastes for VO
      removal.  March 6, 1986.

144.  PEI Associates, Inc.  Field Evaluation of a Sludge Dewatering Unit at
      a Petroleum Refinery.  Volume I, Preliminary Draft.  Prepared for
      U.S. Environmental Protection Agency.  Cincinnati, OH.  August 1981.
      p. 6-28 through 6-29.

145.  Ross, R. D.  The Burning Issue:  Incineration of Hazardous Waste.
      Pollution Engineering.  August 1979.  p. 25-28.

146.  Reed, J. C., and B. L. Moore.  Chapter 12, Ultimate Hazardous Waste
      Disposal by Incineration.  In:  Toxic and Hazardous Waste Disposal,
      Vol. 4.  1980.  p. 163-173.

147.  Bonner, T. A., et al.  Engineering Handbook for Hazardous Waste
      Incineration.  Monsanto Research Corporation.  Dayton, OH.  EPA
      Contract No. 68-03-3025, Work Directive SDM02.  June 1981.  p. 2-26.

148.  Reference 147, p. 2-24.

149.  U.S. EPA/ORD/HWERL.   Innovative Thermal Hazardous Waste Treatment
      Processes.  Cincinnati, OH.  NTIS PB85-192847.  April 1985.  p. 55.

150.  Reference 147, p. 2-30.

151.  Reference 147, p. 2-32.

152.  Sittig, M.  Incineration of Industrial Hazardous Wastes and Sludge.
      Noyes Data Corporation.  Park Ridge, NJ.  1979.

153.  Reference 149.

154.  Reference 130, p. 4-2.

155.  U.S. Environmental Protection Agency.  Petroleum Refinery Enforcement
      Manual.  Research Triangle Park, NC.  January 1980.  p. 4.13-1 ff.

156.  Federal Register.  Volume 50, Number 230.  November 29, 1985.  p. 46-
      48.

157.  Trip report.  Research Triangle Institute.  Visit to Mobil Oil
      Corporation, Joliet Refinery, Joliet, IL.  Prepared for U.S.
      Environmental  Protection Agency.  September 4, 1986.

158.  Telecon.  Wright,  Milton D.,  RTI,  with Weisenborn, Bill, Conoco,  Inc.
      June 20, 1986.  Coking of refinery sludges.

159.  Reference 14,  p.  4.4-6.
                                   4-87

-------
160.  Radian Corporation.  Field Assessment of Air Emissions and Their Con-
      trol at a Refinery Land Treatment Facility: Project Summary.
      Prepared for U.S.  Environmental  Protection Agency.  Cincinnati, OH.
      EPA Contract No. 68-02-3850.   September 1986.   17 p.

161.  Reference 9.

162.  U.S. Environmental Protection Agency.  Fugitive Emission Sources of
      Organic Compounds—Additional Information on Emissions,  Emission
      Reductions,  and Costs.   Office of Air Quality  Planning and Standards.
      Research Triangle  Park, NC.   Publication No. EPA-450/3-82-010.  April
      1982.   p. 2-21.

163.  Reference 162,  p.  4-1  through 4-75.

164.  U.S. Environmental Protection Agency.  VOC Fugitive Emission Sources
      in Synthetic Organic  Chemicals Manufacturing Industry — Background
      Information  for Proposed Standards.   Office of  Air Quality Planning
      and Standards.   Research Triangle Park,  NC.  Publication No. EPA-
      450/3-80-033a.   November 1980.  p.  4-1  through  4-27.

165.  Reference 156.

-------
                          5.0  CONTROL STRATEGIES

     The purpose of this chapter  is to present the general concept of con-
trol strategies and to identify specific example control strategies.  The
example strategies will be used in later chapters to illustrate how impacts
are calculated for each strategy  considered as a possible basis for air
emission standards to be proposed for treatment, storage, and disposal
facilities  (TSDF) under Resource, Conservation, and Recovery Act (RCRA)
Section 3004(n).  In addition, this chapter will discuss specific impacts
to be estimated and will identify the "baseline" against which the nation-
wide impacts of control strategies will be measured.
     As will be shown, the number of potential control strategies for TSDF
is large.  Consequently, this chapter presents only example control  strate-
gies to show how strategies are developed and evaluated.  It should be
emphasized that the examples presented in this chapter will  not necessarily
be the basis for the proposed standards.  The strategies evaluated prior to
proposal of the standards, and the impacts of those strategies, are pre-
sented separately in the project  docket.  The strategy selected as the
basis for the proposed standards, and the factors leading to the selection
of that strategy, will be discussed in the preamble to the proposed stand-
ards published in the Federal Register.
5.1  CONTROL STRATEGY CONCEPT
     As discussed in Chapters 3.0 and 4.0, there are a variety of sources
of organic emissions at TSDF and  several types of emission controls that
can be applied to many of these sources. .The term "control strategy," as
used here,  refers to a unique combination of emission sources, emission
controls,  and action level cutoffs (see Section 5.1.1.3) for applying the
controls.   Different strategies are developed and evaluated to estimate the
impacts of potential regulations.  It is important to recognize, however,
                                    5-1

-------
that although a control strategy Identifies specific emission controls to
be applied to TSDF sources, a regulation written to implement the strategy
may be in terms of performance standards that allow equivalent, or more
effective, controls for compliance.  The possible combinations of sources,
controls, action level cutoffs,  and approaches to selecting strategies for
evaluation from the many combinations possible are discussed in the follow-
ing sections.
5.1.1  Combinations of Emission  Sources, Controls, and Cutoffs
     5.1.1.1  Emission Sources.   TSDF emission source categories are shown
in Table 5-1.  Hazardous waste management processes with similar emission
characteristics and potential  emission controls are grouped together (e.g.,
aerated tanks and aerated impoundments are combined into one source
category).  Landfills, wastepiles,  and land treatment are shown grouped
together under the category of "Land Disposal Units."  This is because
hazardous wastes entering these  types of waste management units are covered
by the land disposal restrictions (LDR), which require treatment of a waste
with best demonstrated available technology (BOAT) before it is placed in
one of these units.  It is presumed that this pretreatment will decrease
the organic air emission potential  of landfills,  wastepiles, and land
treatment.  This presumption will be reviewed after all  requirements under
the land disposal restriction  program are promulgated.
     Surface impoundments are also covered by the LDR, but are exempt from
LDR requirements under certain conditions.  Those impoundments not covered
by the LDR would continue to have significant potential-air emissions and,
therefore, additional  controls for surface impoundments  should be consider-
ed under RCRA Section  3004(n).  More information on the  land disposal
restrictions is presented in a discussion of the baseline in Section 5.3.2.
     Control of organic emissions from equipment leaks and process vents at
TSDF is required by a  separate rulemaking under RCRA Section 3004(n) pro-
posed on February 5,  1987 (52  FR 3748).   The proposed rule requires 95
percent control  of process vent  emissions and a leak detection and repair
program for TSDF sources handling hazardous waste containing greater than
10 percent organics.
                                    5-2

-------
                TABLE 5-1.  TSDF EMISSION SOURCE CATEGORIES
                              Source category
                               Storage tanks

                         Quiescent treatment tanks

               Quiescent impoundments (storage or treatment)

                 Aerated or agitated tanks or impoundments

                               Drum storage

                                  Loading

                             Dumpster storage

                              Waste fixation

                             Equipment leaks9

                              Process vents3

     Land disposal units:  Landfills, wastepiles, and land treatment^5


TSDF = Treatment, storage,  and disposal  facilities.

aThese sources are covered by TSDF air standards for process vents and
 fugitive emissions,  proposed February 5, 1987 (52 FR 3748).
     hazardous wastes placed in these sources must meet pretreatment
 requirements of the land disposal restrictions.
                                    5-3

-------
     5.1.1.2  Controls.  In Chapter 4.0, four basic categories of controls
are mentioned.  Two of those categories — suppression or containment and
organic removal—provide distinctly different bases for the formulation of
control strategies as discussed below.
     As discussed in Chapter 4.0, Section 4.1, suppression or containment
controls prevent or reduce the rate of emissions from processes to which
they are applied.  Given that TSDF are usually comprised of a sequence of
waste management processes, it follows that suppression or containment
controls must be applied to each waste management process in the sequence
to be effective.  Otherwise, the suppression or containment functions only
to transfer the air emissions to a point farther downstream in the facil-
ity.  Ultimately, the wastes must be sent to disposal.  Unless disposal is
by deep well injection or the organics are destroyed (e.g.,  incineration),
the potential for air emissions will continue to exist.
     The organic removal processes (discussed in Chapter 4.0,  Section 4.3)
function as indicated by the name.  Waste streams are treated to remove the
organics from the stream, thus reducing the potential for air emissions
from any process farther downstream.  Organic removal processes could be
placed at almost any point in the waste management sequence, except that
some amount of waste storage and waste preparation (e.g.,  solids separa-
tion) may be necessary upstream of organic removal.  Although the organic
removal process might be placed in front of any of many steps in the waste
management sequence, generally it would be most effective if placed as far
upstream in the waste management sequence as possible, thus reducing the
air emission potential from all the processing steps that follow.  The
organics that are removed from the waste streams may be recovered/recycled,
burned as an energy source, or incinerated to prevent organic air emis-
sions.
     Control strategies based entirely on either suppression controls or
organic removal  are not possible because these controls cannot be 'adapted
to all TSDF emission sources.  However, strategies can be formed that
depend predominantly on one or the other of these control  types.
     5.1.1.3  Cutoffs.  Rather than apply emission controls to waste
management processes without regard to the emission potential  of the wastes
                                    5-4

-------
managed,  a  control  strategy may  include an  action, or  "cutoff,"  level  above
which controls  must  be  applied.  Cutoffs can be  used as  a mechanism  for
prioritizing  emission sources within a waste management  facility for con-
trol and  for  excluding  sources that have low emission  potential.  Cutoffs
can be expressed  in  several formats (e.g.,  in terms of volume throughput,
capacity, or  measured or calculated emissions).  A cutoff format easily
applied and directly related to  emission potential for most TSDF sources is
waste volatile  organic  content.
     A cutoff based  on  waste volatile organic content, as measured by  an
appropriate test  method, could be applied in either of two ways.  One  way
would be  to establish a single cutoff level, applied to waste as it is
received  at a facility  (plant),  that would  dictate what emission controls
are required  on all  waste management processes within  that facility.   A
second way would  be  to  develop separate, source-specific cutoff levels for
each waste management process type.  This second approach would provide
more flexibility  in  applying controls but would  require waste testing  at
more locations.
     5.1.1.4  Possible  Combinations of Sources,   Controls, and Cutoffs.
Table 5-2 presents some potential combinations of TSDF emission sources,
emission  controls, and  cutoff levels to illustrate the large number of
possible  control  strategies.  Although the  number of emission source/con-
trol combinations is fixed, any  level  could be chosen  for a cutoff making
the number of emission  source/control/ cutoff combinations,  and hence the
number of control strategies possible, unlimited.
5.1.2  Approaches to Selecting Control Strategies for  Evaluation
     As demonstrated in the previous section, the number of possible combi-
nations of TSDF emission sources, controls,  and  action level  cutoffs (con-
trol strategies)  is  large.  It is not possible to evaluate every combina-
tion in detail.   Several approaches will be taken to select a limited
number of control strategies for detailed evaluation from among the many
possible strategies.  Some of these are discussed below.
     •     Identify Efficient Strategies.  The mandate of RCRA Section
          3004(n) is to develop standards as necessary to protect
          human health and the environment.   With the many possible
          combinations of emission sources,  controls,   and cutoffs,  it
                                    5-5

-------
          TABLE  5-2.   POTENTIAL  SOURCE/CONTROL/CUTOFF COMBINATIONS
 TSDF  emission  source  category
 Controlsa
Example cutoffs,  (in ppm
    except as noted)"
 Storage  tanks
 Quiescent  treatment tanks
0, 1, 2
0, 1, 2, 3
Quiescent treatment  impoundments   0, 1, 2, 3

Aerated tanks or  impoundments      0, 2, 3

Drum storage                       0, 2

Loading                            0, 1

Dumpster storage                   0, 1

Waste fixation                     0, 2, 3
     500 for level  1
     controls
     1.5 psia for level
     2 controls

     500 for level  1
     controls
     1.5 psia for level
     2 controls

     500

     500

     500

     500

     500

     500
Equipment leaks

Process vents
1 (Required by Pro-
   posed Rule for
2  TSDF Equipment
   Leaks and Vents)
      100,000
Land disposal units0
3  (Required by land
    disposal            NA
    restrictions)
TSDF = Treatment, storage, and disposal  facilities.
NA - Not applicable.

aKey to controls:
     0 = No additional control
     1 = Suppression type controls such  as covers
     2 = Venting air emissions to controls
     3 = VO removal  or pretreatment type controls.

aExample cutoffs are based on volatile organic content of waste in parts
 per million, except as noted for storage tank level  2 controls, which are
 based on vapor pressure in the tank headspace.   The  examples are also
 based on measurement of the volatile organic content as the waste enters
 the TSDF.   Alternatively, different cutoffs could  be set for each waste
 management process  to determine whether controls are required for each
 process.

clncludes certain surface impoundments,  landfills,  wastepiles, and land
 treatment.
                                        5-6

-------
          is likely that many strategies will achieve comparable
          levels of health and environmental protection, but at dif-
          ferent costs.  As a first step in selecting strategies for
          detailed evaluation, the group of strategies that are most
          efficient in meeting health risk and environmental protec-
          tion levels will be identified.  This will eliminate higher
          cost strategies that do not achieve additional health and
          environmental protection for the added cost.  It will also
          avoid overstating the cost of achieving a particular level
          of protection.

     •    Focus on Desired Results.  Another approach will be to focus
          on possible desired results of applying strategies.  For
          example, a desired result might be to reduce individual
          health risk to a certain "target" level.  Similarly,  a
          desired result might be to reduce another measure of health
          risk, such as cancer incidence, to a certain level.

     •    Best Technology.  In contrast to focusing on desired
          results, several strategies will be selected based on the
          best demonstrated control technologies available for TSDF
          emission sources.  Strategies that would require best tech-
          nology for all TSDF sources, as well as strategies that
          would apply best technology to the major emitting TSDF
          sources, will be evaluated.

     •    Maximize Net Benefits.  The monetary value of benefits
          yielded by control strategies, as well as control costs,  can
          be estimated.  Strategies that maximize net benefits can be
          selected for evaluation.

5.2  EXAMPLE CONTROL STRATEGIES
     Two example control strategies were selected for use in later chapters

to illustrate how the impacts of TSDF control strategies are estimated.

These two strategies are shown in Table 5-3.  They are presented for illus-

tration purposes only and will not necessarily be the basis for proposed

standards.
     Example control strategy I is based primarily on the use of add-on

emission controls applied to individual emission sources.  In this example,

if wastes containing more than 500 .ppm of volatile organics (VO) are

handled at a TSDF, quiescent impoundments and aerated tanks or impoundments

would be covered or enclosed and vented to a control device to reduce

organic emissions by 95 percent.  Storage tanks and quiescent treatment

tanks would be covered if the volatile organic content of the stored waste
                                    5-7

-------
               TABLE  5-3.   EXAMPLE TSDF AIR EMISSION CONTROL STRATEGIES3
    Source  category
       Strategy I
      (applled when
     VO > 500 ppm)b
     Strategy  II
    (applied when
   VO > 500 ppm)b
 Storage  tanks

 Quiescent  treatment  tanks

 Quiescent  impoundments
   (storage or  treatment)

 Aerated  or agitated  tanks
   or  impoundments

 Drum  storage


 Loading

 Dumpster storage

 Waste fixation
Cover,  vent to control0

Cover,  vent to control0

Cover,  vent to control


Cover,  vent to control


No control  on existing
  enclosure

Submerged loading

Cover

No control
Cover, vent to control0

Organic removal

Organic removal


Organic removal


Enclosure vented to
  control

Submerged loading

Cover

Organic removal
Equipment  leaks
Process vents
Leak detection  and
  repai r^

95% controld
Leak detection and
  repaird

95% control01
Land disposal sources6
Pretreatment  as  required
  by land disposal
   restrictions
Pretreatment as required
  by land disposal
   restrictions
TSDF   Treatment, storage, and disposal facility.
  VO   Volatile organics.

aThe strategies presented in this table are for use in illustrating how
 impacts associated with controls are calculated in later chapters.  They
 will not necessarily be the basis for'developing standards for proposal.
bCutoff is based on volatile organic content of waste entering the facility
 as measured by a test method, except as noted for covered storage and
 quiescent treatment tanks.   If volatile organic content of waste is above
 cutoff,  then controls are required for all sources as shown.

°Storage and quiescent treatment tanks must be covered when volatile
 organic content of waste is above 500 ppm.  Emissions from covered tanks
 must be vented to a control device  when the waste in the tank has a
 true vapor pressure equal to or greater than 1.5 psia.

dRequired by TSDF air standards for equipment leaks and process vents.
eLandfills,  wastepiles,  and land treatment.
                                      5-8

-------
 is greater  than  500 'ppm.   Covered  storage  and  quiescent  treatment  tanks
 would  be  vented  to an  emission control  device  if  the  vapor  pressure  of the
 waste  in  the  tank exceeds  1.5 psia.   Dumpsters would  be  covered  under
 example strategy  I, and no controls would  be applied  to  drum  storage areas
 or waste  fixation operations.  Loading  operations would  be  controlled by
 using  submerged  loading.   Controls for  process vents  and equipment leaks
 would  be  as required by the proposed  rules  for TSDF'equipment  leaks  and
 process vents, and emissions from  landfills, wastepiles, and  land treatment
 operation would be reduced under the  land  disposal  restrictions.
     Example  strategy  II is based  primarily on the  use of organic removal
 to reduce organic emissions from sources where it is  applicable.  In exam-
 ple strategy  II, organic removal would  be  applied to  wastes entering quies-
 cent treatment tanks,  quiescent impoundments, aerated or agitated tanks or
 impoundments, and waste fixation when the  volatile  organic  content of the
 waste  exceeds 500 ppm.  Controls on the remaining sources would be as in
 example strategy I.
 5.3  IMPACTS  TO BE ESTIMATED FOR CONTROL STRATEGIES
     The  nationwide impacts that will be estimated  for control strategies
 selected  for  detailed  evaluation and  the baseline to  which  the impacts of
 strategies will be compared are discussed below.
 5.3.1  List of Nationwide  Impacts  to  be Estimated
     Implementation of any of the  possible control  strategies for TSDF air
 emissions would have a wide range  of  impacts.  Some impacts, such as  emis-
 sion reduction, reduction  in health risk, and cost  of controls can be esti-
mated,  although with varying degrees  of certainty.  Other impacts of  con-
trol  strategies, such  as ease of implementation, are  difficult to quantify
but are important to consider qualitatively nonetheless.
     For example, organic emissions may contribute  to the formation of
ozone  in the  lower atmosphere (tropospheric ozone)  or the depletion of
ozone  in the  upper atmosphere (stratospheric ozone).  Tropospheric ozone
can adversely affect human health  and the environment.  Stratospheric ozone
is essential  to protecting the earth  from the sun's ultraviolet radiation.
A reduction in nationwide organic  emissions could have a beneficial impact
                                    5-9

-------
on both tropospheric and stratospheric ozone, but because of the complexity
of pollutant transport and ozone formation in the atmosphere, this impact
is not easily quantified.
     To ensure that the consequences of adopting control strategies can be
anticipated as fully as possible, a wide range of nationwide impacts will
be evaluated for control strategies considered as the possible basis for
standards to be proposed under RCRA Section 3004(n).  The impacts to be
evaluated are listed in Table 5-4.
5.3.2  Baseline for Nationwide Impacts Estimates
     The baseline provides a perspective from which the impacts of adopting
control strategies can be evaluated.  Chapter 3 of this document presents
nationwide uncontrolled TSDF organic emissions; i.e., emissions with cur-
rent existing controls and before the adoption of additional nationwide
control requirements.  However,  there are regulations that affect TSDF air
emissions that are, or will  be,  in place when the RCRA 3004(n)  standards
being developed in this rulemaking are promulgated.  The baseline against
which the potential impacts  of the 3004(n) standards are measured should
reflect these other regulatory requirements if they affect TSDF nationwide.
     Federal regulatory requirements that affect TSDF nationwide include
the land disposal restrictions (proposed and promulgated under RCRA Section
3004(m)) and TSDF air standards  for process vents and fugitive emissions
(proposed February 5, 1987,  under RCRA Section 3004(n)).  There are other
Federal requirements applicable  to TSDF,  such as the RCRA Corrective Action
Program (implemented under RCRA  Section 3004(u)),  but these are site-speci-
fic rather than nationwide control requirements.
     Also,  there are standards that apply to TSDF emission sources at the
State level.  However,  these are limited and vary widely from State to
State.  A survey of State programs indicated that 12 States have estab-
lished generic volatile organic  compounds (VOC) standards that might affect
TSDF emission sources.   Thirty States have standards applicable to storage
tanks, 17 States have standards  for terminal loading, and 9 States have
standards for hazardous waste landfills.   For those States that do have
standards,  there are differences in how VOC is defined, making it difficult
to compare requirements from State to State.
                                   5-10

-------
TABLE 5-4.  NATIONWIDE IMPACTS TO BE ESTIMATED FOR CONTROL STRATEGIES


                   Emissions of volatile organics

                          Cancer incidence

                   Maximum individual  cancer risk

                         Cross-media impacts

                       Capital and annual  cost

                         Cost effectiveness

    Cost effectiveness with benefit credit for volatile organics
                         emission reduction

                          Economic impacts

                       Ease of implementation
                              5-11

-------
     After the review of Federal and State standards applicable to TSDF,  it
was concluded that the baseline for the RCRA Section 3004(n) standards
under development should reflect the impacts of the land disposal restric-
tions and the TSDF air standards for process vents and equipment leaks.
Due to the limited applicability and lack of uniformity in State standards,
they should not be included in the baseline.
     More information on how the impacts of the land disposal restrictions
and the TSDF process vent and equipment leak standards will be simulated  in
the baseline is presented below.
     5.3.2.1  Land Disposal Restrictions.  The Hazardous and Solid Haste
Amendments (HSWA) of 1984 mandate a number of actions that EPA must take  to
reduce the threat of hazardous waste to human health and the environment.
These actions include restricting hazardous wastes from land disposal.  The
EPA is currently developing regulations (referred to as land disposal
restrictions) that will require that hazardous wastes be treated to reduce
concentrations of specific chemicals or hazardous properties before the
waste may be placed in a land disposal  unit.  The affected land disposal
units include surface impoundments,  wastepiles,  landfills,  and land treat-
ment operations.  Surface impoundments  used for treatment of hazardous
wastes are exempt from the land disposal restrictions if treatment residues
are removed annually.  The waste treatment technologies required to reduce
chemical concentrations before land disposal are referred to as best demon-
strated available technologies (BOAT).   The restrictions express BOAT as  a
performance standard that requires wastes to be treated before entering a
land disposal unit (to reduce the waste's toxicity or mobility).  The
wastes must be treated to levels that can be achieved by use of the best
technologies commercially available.
     The EPA is proposing and promulgating the land disposal restrictions
in stages,  with the final stage scheduled for promulgation in 1990.  The
first set of land disposal  restrictions, for certain dioxins and solvent-
containing  hazardous wastes,  was promulgated on November 7, 1986
(51 FR 40572).   The second set of restrictions,  the "California list," was
promulgated on  July 8,  1987 (52 FR 25760).
                                   5-12

-------
      The  standards  for  air  emissions  from  TSDF  being  developed  under
 Section 3004(n)  in  the  current  rulemaking  are scheduled  for  promulgation  in
 1990,  when  the  land  disposal  restrictions  should  be fully  in  place.   There-
 fore,  the  land  disposal  restrictions  are considered a part of the  baseline
 against which the potential  impacts of  the 3004(n) standards  are measured.
      To simulate the  impacts  on  TSDF  air emissions from  the  land disposal
 restrictions, the following  presumptions are made concerning  the manner in
 which  TSDF  will  respond  to  the  restrictions:1
      •    All wastes  currently  land-treated  (except organic-containing
          solids) will  be incinerated.  Solids  will be fixated.
      •    All organic liquids and  sludge/slurries that are currently
          sent  to landfills  and  wastepiles will be incinerated  or
          steam-stripped.
      •    All dilute  aqueous  liquids  and aqueous  sludge/slurries are
          fixated and landfilled.
      •    All surface impoundments will be converted  to  uncovered
          tanks.
      Because the complete requirements  of  the land disposal  restrictions
 have  not yet been promulgated,  the above presumptions may not be valid for
 all TSDF.   However,  on  a nationwide basis, it is  likely  that  they  represent
 the general or  average  response  of the  hazardous  waste management  industry.
 Therefore,  it is reasonable  to  include  them  in  the baseline  against which
 the nationwide  impacts  of potential air standards under  RCRA
 Section 3004(n) are  evaluated..
      5.3.2.2  TSDF Air  Standards for  Equipment  Leaks  and Process Vent
 Control.   On February 5, 1987,  the EPA  proposed RCRA  air emission  standards
 for volatile organics control at TSDF.  These standards  are  intended to be
 an "accelerated" portion of  the  RCRA  3004(n) air  standards.   The proposed
 standards apply to new  and existing facilities  and pertain to both process
 and equipment leak air emissions.  Sources of process emissions include
 process condenser vents, and distillate receivers, surge control vessels,
 product separators,  and hot wells, if process emissions  are  vented through
 these vessels.   Equipment leak  sources  include  pumps,  valves, pressure-
 relief devices,  compressors, open-ended lines,   and sampling  connections.
These standards will apply to hazardous wastes  containing greater  than
                                   5-13

-------
10 percent organics.  Specific requirements include reduction of process
vent emissions to less than 1.4 kg/h (3 Ib/h)  and 2.8 Mg/yr (3.1 tons/yr)
or reduction of facility process vent emissions by 95 percent; and a leak
detection and repair program for equipment leaks.
     To simulate the impact of the accelerated rule on the baseline,  all
process vents and equipment leak emissions at  facilities handling wastes
containing greater than 10 percent organics will  be controlled as part of
the baseline.
5.4  REFERENCE
1.   Industrial  Economics,  Inc.,  and ICF Incorporated.  Regulatory Analysis
     of Proposed Restrictions  on Land Disposal  of Hazardous Wastes.   Pre-
     pared for U.S.  Environmental  Protection Agency.   Washington, DC.
     December 1985.   p. 4-20 through 4-22  and  exhibits.
                                  5-14

-------
          6.0  NATIONAL ORGANIC  EMISSIONS AND HEALTH RI.SK  IMPACTS

     The nationwide organic air  emission and health risk impacts associated
with each of the two treatment,  storage, and disposal facility  (TSDF)
example control strategies identified  in Chapter 5.0 are discussed in this
chapter.  The primary emphasis of Chapter 6.0 is to present the approach
for assessing these impacts by using the example control strategies as a
guide for discussion and presentation.  Both beneficial and adverse
environmental impacts are assessed.  Table 6-1 summarizes  these nationwide
impacts; presented are the nationwide  estimates of TSDF organic emissions,
cancer incidence, and maximum  lifetime  risk for the two example control
strategies as well as the uncontrolled  and baseline scenarios.  Comparisons
to the baseline situation are  also made to provide a relative measure of
the effectiveness of or degree of control required under the two example
strategies.  More detailed information  on the emissions and health risks
associated with implementation of the  example control strategies are
provided in Sections 6.1 and 6.2.  The  nationwide emission reduction for
each example control strategy  is tabulated in Section 6.1.  Impacts on
human health, assessed as cancer incidence and maximum lifetime risk of
contracting cancer, are presented in Section 6.2.  Impacts on water
quality, solid waste,  energy,  and other environmental concerns are under
development and will be presented in Section 3.6 at a later date.
6.1  ORGANIC EM-ISSION IMPACTS
     This section presents the nationwide impacts on TSDF  organic air emis-
sions for two example control  strategies and includes a description of how
emissions under the uncontrolled, baseline, and controlled scenarios were
estimated.   A tabular presentation of the estimated emission reductions
from affected waste management units is also included.  As discussed in
Chapter 5.0,  the baseline case assumes that Resource Conservation and
                                    6-1

-------
en
i
ro
                           TABLE 6-1.   SUMMARY OF NATIONWIDE ORGANIC  EMISSIONS  AND  HEALTH  RISK  IMPACTS FOR
                                     UNCONTROLLED,  BASELINE, AND  TWO  EXAMPLE CONTROL  STRATEGIES3
Emission reduction Cancer
Emissions, from baseline, Percent incidence, c
Contro 1 case 103 Mg/yr^ 103 Mg/yr reduction, % incidence/yr
Uncontrolled 1,810 — — 130
Baseline 1,835 — — 135
Example control
strategy Id 364 1,471 80.2 21
Example control
strategy IIe 349 1,486 81.0 15
Incidence reduction Maximum
from baseline lifetime
incidence/yr risk, MLR
1 x 10-2
1 x 10-2
114 4 x 10-4
120 5 x 10~4
TSDF = Treatment, storage, and disposal facility.
  — = Not applicable.

aWith TSDF proposed air standards (for process vents and equipment leaks) and land disposal restrictions in place.
^Refer to Table 6-2 for nationwide emissions by source.
cRefer to Table 6-3 for nationwide cancer incidence by source.
^Example control strategy I applies to wastes containing greater than 500 ppm of organics.  It entails covers and
 controls for tanks and impoundments, submerged  loading of drums, and covers for dumpsters.  For covered storage and
 quiescent treatment tanks, venting to a control device is  required if the vapor pressure of the waste in the tank
 exceeds 1.5 psi a.
eExample control strategy II applies to wastes containing greater than 500 ppm of organics.  It entails introducing
 organics removal processes before treatment tanks, storage or treatment impoundments, and waste fixation processes;
 covers and controls for storage tanks as in example strategy I; enclosure and control of drum storage areas; submerged
 loading of drums;  and covers for dumpsters.

-------
 Recovery  Act  (RCRA)  land  disposal  restrictions and the  1987 proposed TSDF
 air  emission  standards  (52  FR  3748)  are  in effect.   Example strategies  I
 and  II  represent  two  potential  "controlled" cases.   Data  are presented  for
 these three scenarios to  allow comparison of the  impacts  of the example
 control strategies  in relation  to  current and 1987 proposed standards on
 TSDF air  emissions  nationwide.
     Chapter  3.0  of this  document,  "Industry Description  and Air Emis-
 sions," provides  nationwide  estimates of uncontrolled emissions by TSDF
 management process.   Nationwide emissions were computed using the computer-
 ized Source Assessment Model  (SAM)  by first identifying all process source
 categories listed  in  the  Industry  Profile (described in Appendix D.2.1).
 Once these categories were  identified, their emissions were calculated by
 multiplying the organic quantity of  each waste stream by  an emission factor
 specific  to the particular management process and the wastes being
 processed  (see Appendix D,  Section  D.2.4.1).  Emissions per process per
 TSDF then were summed to  yield  a nationwide uncontrolled  emission estimate.
     To calculate  the quantity  of  emissions reduced by applying organic
 emission  controls,  the control  technologies described in  Chapter 4.0 were
 applied to the appropriate waste management processes (source category) as
 required  by one of  two example  control strategies:
     •    Example  control strategy  I applies to wastes containing
          greater  than 500 ,ppm  of  organics.  It entails covers and
          controls  for tanks and impoundments, submerged  loading of
          drums,  and  covers  for dumpsters.  For covered storage and
          quiescent treatment  tanks, venting to a control device is
          required  only if the  vapor pressure of the waste in the tank
          exceeds  1.5 psia.
     •    Example  control strategy  II also applies to wastes con-
          taining  greater than  500  ppm of organics.  It entails
          introducing organic  removal processes before treatment
          tanks,  storage  or  treatment impoundments, and waste fixation
          processes;  covers  and controls'for storage tanks (with
          control  devices for  covered tanks required only if the vapor
          pressure of the stored waste exceeds 1.5 psia); control of
          enclosed drum storage areas; submerged loading  of drums; and
          covers for dumpsters.
     The magnitude of nationwide organic emissions associated with each
example strategy was calculated using the SAM.  In short, this consisted of
adjusting the uncontrolled and  baseline emissions by the  control efficiency
of the control technology required under each particular  strategy, for each
                                    6-3

-------
TSDF process stream at each facility nationwide.  Summation of results
provides an estimate of emissions per control strategy.  Table 6-2 presents
the nationwide results by emission source and the emission reductions
resulting from implementation of each control strategy by source category.
Because the specific design and operating characteristics of each waste
management process are not widely available, nationwide distributions of
process design and operating parameters were used in estimating TSDF emis-
sions.  Therefore, it is appropriate that nationwide TSDF emissions and
impacts are used in the comparison of the various control options and
strategies.  As noted in Chapter 3.0, the estimation of TSDF emissions in
this document involved the use of the TSDF air emission models as presented
in the March 1987 draft of the air emission models report rather than the
December 1987 version of the report.
     Nationwide uncontrolled emissions were estimated at about 1.81 million
Mg/yr; baseline emissions were 1.83 million Mg/yr.  Storage tanks are the
largest uncontrolled emitters nationwide under these two scenarios.
Process vents from organic removal devices have zero baseline emissions
because these processes are not considered to be operating uncontrolled;
that is, for process vents it is assumed that emissions are controlled
under Federal regulations currently in effect.
     Emission reductions from baseline for the two strategies range from
1.486 million Mg/yr (strategy II) to 1.471 million Mg/yr (strategy I).
Control devices for tank storage yield the highest emission reductions in
strategies I and II (99.7 and 99.3 percent, respectively).  Some sources
such as waste fixation processes show an increase in emissions when a
control strategy is applied.  These increases occur because (1) emissions
are suppressed from upstream controlled sources (i.e., the waste stream
retains the organics that would have previously been emitted, which results
in an increase in organics at the source of interest), and (2) when
controls are applied,  new emission sources are created such as pumps and
valves (i.e., equipment leaks).
6.2  HUMAN HEALTH RISKS
     Health risks posed by exposure to TSDF air emissions are presented in
this section in three forms:  annual cancer incidence  (incidents per year
nationwide resulting from exposure to TSDF air emissions), maximum lifetime
risk (the highest risk of contracting cancer that any  individual could
                                    6-4

-------
                                                      TABLE 8-2.  NATIONWIDE TSDF EMISSIONS AND EMISSION REDUCTION FOR THE UNCONTROLLED, BASELINE,
                                                                                   AND TWO EXAMPLE CONTROL STRATEGIES8
CT)
 I
cn
Uncon tro 1 1 ed
Source category
Drum storage
Dumpster storage
Storage tanks
Quiescent surface
i mpoundments
tanks
Aerated tanks and
surface impoundments
Was topi ies
Landf i 1 Is
Waste fixation process
Inc i nera t Ion'
Land treatment
Other treatment9
Spi 1 Is
Load i ng
Equipment leaks
TOTAL*1
emissions,
103 Mg/yr
0.
78
766
209
48
BIB
0.
40
2
0
73
0
0
6
80
1,810
19




,13

.1
,89


.43
.8


See notes at end of table.

Basel in*
emi ss i ons , k
103 Mg/yr
0.19
78
768
208
48
BIB
0.033
2.1
177
1 .1
0
0
0.42
6.8
40.3
1,836

Emi ss i ons ,
103 Mg/yr
0.19
2.1
3.9
42
16
27
0.033
3.8
220
1.1
0
0
0.43
B.I
43
364
Example control stri
Emi ss i on
reduct i on
from base 1 i ne ,
103 Mg/yr
0
76
764
167
33
488
0
-1.7"
-44"
0
0
0
-0.01"
1.7
-2.7"
1,471
tegy Ic
Percent
reduction, !?
0
97
99
80
69
96
0
-
-
0
0
0
-
26
-
80.2
Examp 10
Emi ss i ons ,
103 Mg/yr
0.013
2.1
7.2
64
18
196
0.004
0.60
9.8
0.96
0
3.0
0.49
3.9
44
349
control strategy
Emi ss i on
reduct ion
from basel Ine,
103 Mg/yr
0.18
76
761
146
30
320
0.029
l.B
167
-0.15°
0
-30
-0.07°
2.9
-3.7°
1,486
Il'd
Percent
reduct i on ,
96
97
99
69
63
62
88
71
94
-
0
-
-
43
-
81.0
                                                                                                                                                                        (conti nued)

-------
                                                                               TABLE 6-2 (continued)
           TSDF = Treatment,  storage,  and disposal facility.

              - = No reduction achieved.

           aEstimated values  for each  source category should not be used or vi ewed independently because some control  technologies may suppress
            emi ss i ons or reduce emi ssions (e.g., organic removaI  processes) that will  impact sources downstream.

           ^Wt th TSDF proposed air standards (for process vents and equi pment leaks)  and land disposal  restrictions in place.

           cExample control  strategy I applies to wastes containing greater than 600 ppm of organ i cs.   It entails covers and controls for tanks and
            Impoundments,  submerged  Ioad1ng of drums, and covers for dumpsters.   For covered storage and qu iescent treatment tanks,  venting to a controI
            dev i ce is  requ i red if the vapor pressure of the waste in the tank exceeds 1.5 ps! a.

           ^ExampIe controI  strategy II applies to wastes containing greater than 500 ppm of organ i cs.   It entails introducing organics removaI processes
            before treatment  tanks,  storage or treatment impoundments, and waste fixation processes; covers and controls for storage tanks as in
            examp Ie strategy  I; enclosure and control of drum storage areas; submerged Ioad i ng of drums; and covers for dumpsters.

           eNegat i ve numbers  mean an Increase \n emi ssions.   Th is is due to em!ss i ons suppressed from  upstream controI led sources,  increas i ng organ i cs
            loading at the source of interest, and the creat i on of new emi ss i on sources such as pumps  and valves (i.e.,  equi pment Ieaks)  when controIs
            are appI i ed.

           "Uncontrolled incinerator emissions include emi ssions from wastes that are routinely incinerated with stack exhaust gas emission controls.
 I           These sources  are currently regulated under 40 CFR 264 Subpart 0.  The uncontrolled emi ssion scenario does not include wastes that are or
Q>,          wouId be i nc i nerated as  a  result of implementi ng the RCRA land disposal restr ictions.  The baseline and two examp Ie strategies do, however,
            account for the  incinerator emi ssion resulting from the land disposal restrictions.  The emi ssion scenarios are explained in  Chapter 5.0.

           90ther treatment  includes processes such as stream stripping that are typically used to  remove organics from wastes.  For the  uncontrolled
            em!ssion case,  these emi ssions are built into the tank treatment category because of similarities in emi ssion characteristics.

           "The sum of each  coIumn of  data does not equal the total value because of rounding.

-------
 have  from  exposure  to  TSDF  emissions),  and  noncancer  health  effects  (from
 acute  and  chronic exposures  to  noncarcinogenic  chemical  emissions  from
 TSDF).   Annual  cancer  incidence and  maximum lifetime  risk  are  used  as an
 index  to quantify health  impacts  for the  three  control  cases:
 (1) uncontrolled air emissions,  (2)  baseline  air  emissions,  and  (3)  control
 under  two  example strategies.
     Detailed discussion  on  the development of  the  health  effects  data
 presented  here  are  found  in  Appendixes  E  and  J.   In general, the methodol-
 ogy consists of four major  components:  estimation  of the  annual average
 concentration patterns  of TSDF  organic  air  emissions  in  the  region  sur-
 rounding each facility, estimation of the population  associated with each
 computed concentration, estimation of exposures computed by  summing  the
 products of the concentrations  and associated populations, and, finally,
 estimation of annual incidence  and maximum  lifetime risk,  which are
 obtained from exposure  and  TSDF emission  potency  data.
 6.2.1  Annual Cancer Incidence
     For the estimates  of TSDF  incidence, the Human Exposure Model  (HEM),
 which  uses a basic  EPA  dispersion algorithm, was  used to generate organic
 emission concentration  patterns.  The TSDF  Industry Profile  (see Appendix
 D.2.1) was accessed to  identify facility  locations  for  population pattern
 estimation within the  HEM using 1980 census  population  distributions.  The
 HEM was  run for each TSDF using a fixed unit  risk factor and a facility
 organic  emission rate;  as such,  the  HEM site-specific incidence results can
 be adjusted by  the  annual facility emissions  generated  from  the SAM  and the
 appropriate TSDF unit  risk  factor to give facility-specific  estimates for
 the control strategy under  consideration.   The  incidence results therefore
 reflect  the level of emissions  resulting  from a particular emission
 scenario or control strategy.
     As  shown in Table  6-3,  incidence estimates indicate that  an uncon-
 trolled  TSDF industry would  lead to  130 cancer  incidents per year nation-
 wide;  the baseline TSDF industry case would  lead  to 135  incidents.   Example
 control  strategy I reduces  the  estimated  number of  cancer  incidences some
 84 percent from 135 in  the  baseline  case  to  21  per year.   Example control
 strategy II reduces the estimated number  of  cancer  incidences  about 89 per-
 cent from 135 incidences  in  the baseline  to  about 15 cancer  incidences per
year nationwide.

                                     6-7

-------
                          TABLE 6-3.  NATIONWIDE CANCER INCIDENCE FROM TSDF EMISSIONS BY SOURCE CATEGORY3.b
                                                 (Number of Cancer Incidences/Year)
cr>
i
03
Cancer incidence due to ...
Source category
Drum storage
Dumpster storage
Storage tanks
Quiescent surface impoundments
Quiescent treatment tanks
Aerated tanks and surface
impoundments
Was top i les
Landf i 1 Is
Waste fixation
Incinerati on'
Land treatment
.Other treatments
Spi 1 Is
Load i ng
Equipment leaks
TOTAL
Uncontro 1 led
emi ssions
0.016
6.8
72.2
8.6
3.3
30.3
0.021
1.9
0.13
0.032
0.66
0
0.041
0.62
5.8
130
Basel ine
emi ss ionsc
0.016
6.8
72.4
8.7
3.3
30.4
0.003
0.12
9.8
0.043
0
0
0.040
0.62
2.8
135
Cancer incidence after . . .
Example control strategies
-j3—
0.016
0.12
0.28
0.90
1.3
1.6
0.003
0.063
13.0
0.043
0
0
0.041
0.44
3.0
21
IIe
0.001
0.12
0.43
2.2
1.0
7.5
0
0 . 0003
0.53
0.043
0
0.11
0.048
0.35
3.0
15
             TSDF = Treatment, storage, and disposal facility.

             aThis table shows nationwide cancer incidence by emission source.  Each of the four categories of incidence
              listed above is described in detail in Chapter 5.0.  Development of incidence estimates is presented in
              Appendix E.  This table provides only example control strategy data.  It is shown only for the purpose of
              demonstrating the estimation of impacts of control strategies.

             ''Estimated values for each source category should not be used or viewefj independently because some control
              technologies may suppress emissions or reduce emissions (e.g., organic removal processes)  that will impact
              sources downstream.

                                                                                                             (conti nued)

-------
                                                 TABLE 6-3 (continued)
CTl
cWith TSDF proposed air standards (for process vents and equipment  leaks) and  land disposal restrictions  in
 pI ace.

^Example control strategy I applies to wastes containing greater than 500 ppm of organics.  It entails
 covers and controls for tanks and impoundments, submerged  loading  of drums, and covers for dumpsters.
 For covered storage and quiescent treatment tanks, venting to a control device  is required if the
 vapor pressure of the waste in the tank exceeds 1.5 psia.

eExample control strategy II applies to wastes containing greater than 500 ppm of organics.  It entails
 introducing organics removal processes before treatment tanks, storage or treatment  impoundments, and waste
 fixation processes; covers and controls for storage tanks as  in example strategy I; enclosure and control
 of drum storage areas; submerged loading of drums; and covers for  dumpsters.

'Uncontrolled incinerator emissions include emissions from wastes that are routinely  incinerated with stack
 exhaust gas emission controls.  These sources are currently regulated under 40 CFR 264 Subpart 0.  The
 uncontrolled emission scenario does not include wastes that are or would be incinerated as a result of
 implementing the RCRA  land disposal restrictions.  The baseline and two example strategies do, however,
 account for the  incinerator. emission resulting from the  land  disposal restrictions.  The emission scenarios
 are explained  in Chapter 5.0.

SOther treatment  includes processes such as stream stripping that are typically used to remove organics from
 wastes.  For the uncontrolled emission case, these emissions  are built  into the tank treatment category
 because of similarities  in emission characteristics.

-------
6.2.2  Maximum Lifetime Risk
     Maximum lifetime risk (MLR) represents "individual" risk as opposed to
the "aggregate" risk in the total nationwide cancer incidence and is
intended to reflect the Nation's most exposed individual's chance of
getting cancer if exposed continuously for 70 years to the highest annual
average ambient concentration around a TSDF.  As such, MLR reflects the
highest risk that any person would have from exposure to TSDF emissions.
MLR is calculated as a function of ambient organic concentration and the
composite unit risk factor for TSDF organic emissions.  For TSDF MLR
estimates, the Industrial Source Complex Long-Term Model (ISCLT), a state-
of-the-art air quality dispersion model, was used to generate the maximum
annual average ambient organic concentration estimates (see Appendix J for
a description of the model).  In order to provide a more comprehensive
analysis of maximum ambient concentrations, two TSDF were selected for
detailed, rigorous analysis in making MLR estimtes.  The two facilities
were selected on the basis of their estimated emissions and the TSDF
management processes utilized at the facilities.  The design and operating
parameters and wastes managed at these two facilities were used in conjunc-
tion with the local meteorological conditions (standard climatological
frequency of occurrence summaries) to estimate dispersion of emissions from
each source at each facility on an annual basis.  Multiplying the maximum
annual average ambient concentration by the composite unit risk factor
yields the maximum risk,  given that someone is predicted to reside at that
location.  The unit risks from the various individual dispersed carcinogens
are represented by a composite unit risk factor for TSDF organic emissions.
Pertinent information on  the selected TSDF and unit risks is presented in
Appendix J and Appendix E, respectively.
     The results of the MLR calculations, shown in Table 6-4, indicate that
the probability of contracting cancer is 1 x 10~2 for the baseline TSDF
industry.  For example control strategies I and II, these risks are
4 x 10~4 and 5 x 10"^,  respectively.  For both example strategies, surface
impoundments are the major sources contributing to the maximum ambient air
concentrations associated with the MLR values.
                                   6-10

-------
        TABLE 6-4.  MAXIMUM LIFETIME RISKS FROM
                    TSDF EMISSIONS3
Control  scenario
                        Maximum
                    concentration,
                                         Maximum risk^
Uncontrol led
Basel ine
Strategy Ic
Strategy IIC
1,700
1,700
50
60
1 x ID'2
1 x ID'2
4 x lO'4
5 x ID'4
aThis table shows the cancer risk of the individual  in
 the United States most exposed to TSDF emissions over a
 70-year period.  Risk is presented for four control
 scenarios, which are described in detail  in Chapter
 5.0.  Development of risk data is presented in Appen-
 dixes E and J.

^Based on a composite risk factor of 8.6 x 10"6
cBased on 500 ppm volatile organic cutoff and an addi-
 tional  cutoff of 1.5 psia for storage and quiescent
 treatment tanks (to determine level  of control
 requi red) .
                         6-11

-------
6.2.3  Noncancer Health Effects Assessment-.-Acute and Chronic Exposures
     A screening analysis of the potential adverse noncancer health effects
associated with acute and chronic exposure to individual waste constituents
emitted from the two selected TSDF was based on a comparison of relevant
health data to the highest short-term (i.e., 15-min, 1-h, 3-h, and 24-h) or
long-term (i.e., annual) modeled ambient concentrations for chemicals at
each facility (see Appendix E).  Modeled concentrations were estimated from
the Industrial Source Complex-Short Term (ISCST) Model.  Detailed
information on this model and on modeled ambient concentrations of
constituents at each facility is provided in Appendix J.
     Results of this analysis indicate that adverse noncancer health
effects are unlikely to be associated with acute or chronic exposure to the
given ambient concentrations of individual chemicals at these two TSDF.
Modeled short-term and long-term ambient concentrations were in most cases
at least three orders of magnitude below health effects levels of concern.
It should be noted that the health data base for many chemicals was
limited, particularly for short-term exposures.  The conclusions reached in
this analysis should be considered in the context of the limitations of the
health data, the uncertainties associated with the characterization of
wastes at the two facilities,  and the assumptions used in estimating
emissions,  ambient concentrations,  and the potential for human exposure.
6.3  OTHER ENVIRONMENTAL IMPACTS
(Data to support this section are currently under development.)
                                   6-12

-------
                         7.0   NATIONAL  CONTROL COSTS

     The purpose of this  chapter  is  to present the general methodology used
to estimate nationwide costs  of adopting  each of the two example control
strategies described  in  Chapter 5.0  as the basis for regulation of air
emissions from  hazardous  waste treatment, storage, and disposal facilities
(TSDF).  Estimated nationwide total  capital  investment (TCI; i.e., equip-
ment purchase and direct  and  indirect  installation costs) and total annual
costs  (TAC; i.e., costs  of operating control technologies minus any energy
or materials credits) are provided  in  a subsequent section of this chapter.
In addition, the cost per unit of waste throughput for the control technol-
ogies  identified in Chapter 5.0 as  part of the example control  strategies
is discussed and listed,  and  a general  explanation of the methodology used
to derive those unit  costs is presented in this chapter.   Supporting data
are provided in Appendixes H  and  I  to  this document, and other references
to cost information are  listed in those appendixes.
     Development of costs requires  the presumption of a baseline level of
emissions and emission control from which the control costs can be calcu-
lated.  The baseline  used for this  effort is described in Chapter 5.0,
Section 5.1, of this  document.  Costs  to  implement the example control
strategies are  provided  to permit a  comparison of the resources that would
be expended to  reduce air emissions  from  TSDF using different combinations
of controls and sources.
7.1  CONTROL COSTS DEVELOPMENT
     Estimation of the nationwide costs of adopting a control strategy
begins with estimation of the control  costs  for individual waste management
units within a TSDF.  Ideally, information about the design and operating
characteristics (such as  surface area  and retention time for impoundments)
of each waste management  unit would be  available to permit accurate
                                    7-1

-------
estimates of control  costs for that unit.  Information at that level of
detail  is not available for each unit at each facility; generally, only
waste throughput is known.  For this reason,  model units were developed.
Rationale for the development of model  units  is given in Chapter 3.0,
Section 3.2.1 (as relates to emission estimation)  and Section C.2.  Control
cost estimates were developed for each  of the model  units as well as
organic removal  processes and hazardous waste incineration (results are
shown in Section C.2.3).  The methodology for developing control  costs for
the model units and the other processes is described partially in Section
7.1.1 and in detail in Appendixes H and I.
     To obtain nationwide costs from model unit costs requires a method of
assigning a model unit cost to each waste management unit in each facility
and then computing the sum.  Given that only  TSDF waste management unit
throughput is known,  the assignment of one of the defined model units to
represent each TSDF waste management unit is  not possible.  Therefore, a
weighted average model unit control cost—in  essence, "national average
model unit" control cost—was derived for each control applied to each TSDF
waste management unit.  These control costs,  divided by the model unit
throughput, provide cost factors that are used to generate control cost
estimates for each TSDF facility.  The discussions of weighted average
model unit control costs and control costs as a function of throughput are
given in Sections 7.1.2.2 and 7.1.2.1,  respectively.
7.1.1  Methodology for Model Units, Organic Removal, and Waste
       Incineration Control Costs
     To estimate the nationwide cost impacts  of implementing the two exam-
ple control strategies presented in Chapter 5.0, the estimated total capi-
tal investment and total annual costs were developed for each of the vari-
ous control technologies applied in the control strategies.  (A general
discussion of these control technologies is contained in Chapter 4.0.)  The
control strategies describe the control technology for each source category
in general terms, such as cover and vent to control  device.  The specific
control technologies assumed to be applied to each source category for each
example strategy are defined in Chapter 5.0,  Section 5.2.
     A standardized cost estimating approach  was developed for add-on and
suppression-type control devices and organic  removal processes based on an
                                    7-2

-------
 EPA  cost  manual^  and  a  series  of  articles  by  Vatavuk  and  Neveril.2-7   These
 sources  identified  the  total capital  investment,  annual operating  costs
 (costs of operating control  technologies minus  capital  recovery  and energy
 credits),  and  the total  annual  costs  (i.e.,  annualized  costs)  as the  key
 elements  of  a  cost  estimate.   Costs of  incineration control  technologies
 were  developed  using  information  provided  by  the  Office of  Solid Waste
 (OSW)-8   These  approaches  are  discussed  in greater detail in Appendixes H
 and  I.
      For  each  control technology  applied in  a control strategy,  a  detailed
 cost  estimate  was developed.   The detailed cost estimate  consisted of three
 standard  cost  tables.   The first  of the  three cost tables lists  the major
 equipment items associated with the control  technology.   The second table
 lists any auxiliary equipment  required,  instrumentation,  sales tax and
 freight,  plus  direct  and  indirect installation  charges.   These first two
 tables are used to  calculate total capital  investment.  The  third  table
 lists the direct  operating costs,  indirect operating  costs,  and  energy
 credits used to calculate  total annual costs.   Examples of these three
 tables are presented  in  Appendix  H.
      The  purchase cost,  material  of construction, and size of each major
 equipment item were obtained from vendor data,  engineering handbooks,  the
 literature,  and currently  operating commercial  facilities.   (Such  sources
 are referenced  in Appendixes H  and I.)  The  sum of the costs for the major
 equipment  items is  equal to the base  equipment  cost (BEC).
      Using the base equipment  cost, the purchased equipment cost (PEC) for
 the control  technology  is  computed.   Direct  and indirect  installation
 charges for  each  control technology are factored  directly from the pur-
 chased equipment  cost.   For this  analysis, the  direct and indirect instal-
 lation factors are  based on information obtained  from vendors,  other cost
 estimates, data summarized  in  References 2 through 7,  and engineering judg-
ment based on typical TSDF wastes  and operating practices.  The costs for
 site preparation  and buildings were based  on  vendor information and
construction cost reference sources.9  The sum of the purchased equipment
cost,  direct installation  charges, and indirect installation charges are
equal  to the total  capital investment (TCI).
                                    7-3

-------
     The sum of direct and indirect operating expenses less capital  recov-
ery and energy credits is equal  to the annual operating costs.   The total
annual cost is equal to the direct plus indirect operating costs less any
energy credits.
     To illustrate this cost approach for add-on control  devices and for
organic removal processes, Appendix H gives a detailed cost analysis for
application of a fixed roof to an open tank vented to a fixed-bed carbon
adsorption system, and Appendix  I shows the analysis for steam stripping of
a dilute aqueous waste.  Similar analyses were performed for the other
control technologies used in the example strategy definitions in Chapter
5.0 and are contained in the document that provides the rationale of the
design, costing, and material and energy balances for TSDF control
options.10  In addition, this document provides flow diagrams for organic
removal processes, sample calculations, and other details of how each con-
trol was costed.
     The costs for waste incineration were not developed from basics, but
made use of a detailed study of  hazardous waste incineration performed for
the OSW.I*  In the OSW study, incineration costs for a variety of wastes
were estimated.  Several of the  OSW incineration estimates for wastes with
compositions similar to model wastes described in Appendix C,
Section C.2.2, were selected as  the basis for the TSDF incineration costs.
This cost-estimating effort is described in more detail in Appendix I.
7.1.2   Derivation of Unit Costs to Estimate Nationwide Costs of
        Example Control Strategies
     The estimation of nationwide costs of the example control  strategies
makes use of a TSDF Industry Profile data base (assembled to aid in this
effort and described in Appendix D) and the emission control costs for
individual source category/emission control combinations whose development
is discussed in Appendixes H and I.  The Industry Profile gives the waste
throughput data used to assign throughputs to each TSDF waste management
unit.  To facilitate the use of  these two information sources,  the total
capital investment and annual operating cost for each of the model unit
cost estimates were divided by the throughput of the model waste management
unit (emission source) to obtain a cost (both total capital  investment and
                                    7-4

-------
 annual  operating  cost)  per  unit  of  waste  throughput.   The  following  para-
 graphs  discuss  the  development of the  unit  cost  factors.
      7-1.2.1  Costs  as  a  Function of Throughput  (Unit  Cost  Factors).  As
 part  of  the effort  to characterize  the variety of  TSDF operating practices,
 model TSDF waste  management  units were defined.  The main  purposes of the
 model units are to  evaluate  uncontrolled  emissions  from waste management
 processes, assess the reduction  in  air emissions when  emission controls are
 applied,  and estimate the costs  of  applying  controls.   Model units were
 defined  for TSDF  storage, transfer  and handling, treatment,  and disposal
 operations.  The  model  units  cover  a range  of waste management unit  sizes
 (e.g., throughput,  surface  area, and tank volumes)  and other characteris-
 tics  that may impact air  emissions.  The  entire  set of model units is
 presented in Appendix C of  this  document.   The approach to  developing con-
 trol  costs discussed in Section  7.1.1  was applied  to each of the model
 units listed in Appendix  C  for each of the  individual  unit  emission  con-
 trols.
      The  next step  toward generating the  control costs  on a nationwide
 basis was to convert the  costs of controls  for the model units, organic
 removal  processes,  and  incineration operations to  a cost per unit of waste
 throughput; i.e., the costs  of controlling  emissions from each model  unit
 or other  process  were divided by the annual  waste  throughput of the  model
 waste management  unit to which the  control was applied.  These factors
 (referred to as unit cost factors), when  multiplied by  the waste throughput
 for a particular  waste management unit, yield an estimate of the cost of
 air emission controls for that unit.
      7.1.2.2  Development of  Weighted  Cost  Factors.  Data contained  in the
 TSDF  Industry Profile (described in Appendix D) are used to estimate annual
waste throughput  for each type of waste management unit at each TSDF.  The
 Industry  Profile, however, does  not yield the exact size of each management
 unit,  e.g.,  758 m^ of tank storage  could  be  ten 75.8-m3 tanks or one 758-m3
tank.   Because there are economies  of  scale  associated with emission
control  costs,  tjie total control  costs might be substantially different for
these two examples.  To compensate  for the lack of facility-specific unit
size information,  weighted unit cost factors were developed that account
                                    7-5

-------
for the national size distribution of TSDF waste management units.
Statistics on the national distribution of waste management unit sizes were
used to weight the emission control costs for each model unit size defined
in Appendix C.1^  This approach yields an approximation of the effects of
economies of scale for the nationwide cost estimates.
     The cost weighting procedure was not used for organic removal and
hazardous waste incineration processes.  These processes would typically
not be used to handle waste streams with a low throughput.  The throughputs
used for estimating the costs of these processes were selected to be within
the range of existing on-site operations for air and steam strippers and
commercial-scale operations for batch distillation, thin-film evaporation,
and incineration.  Consequently, single point unit cost estimates were used
for these processes at relatively large unit sizes where changes in
throughput would not yield major changes in economies of scale.
     Table 7-1 lists the unit cost factors used to estimate nationwide
total capital investment and annual operating costs for each of the emis-
sion controls specified in the example control strategies.
7.2  SUMMARY. OF NATIONWIDE CONTROL COSTS FOR CONTROL STRATEGIES
     This section presents tabular summaries of the estimated nationwide
total capital investment and total annual costs (annualized capital cost
plus annual operating costs) for the two example control strategies
described in Chapter 5.0.  Separate cost estimates are given for each TSDF
source category.
     The nationwide cost estimates were obtained by multiplying the quan-
tity of wastes managed in each TSDF waste management unit (obtained from
the TSDF Industry Profile) by the unit cost factors listed in Table 7-1.
The estimated costs for each TSDF were summed to produce national totals.
Table 7-2 lists the estimated nationwide total capital investment and the
estimated nationwide total annual cost, respectively, for each of the exam-
ple control strategies.
     For the example strategies, the estimated total capital investment for
control  strategy I is $1.084 billion and for control strategy II is $7.19
billion.   The estimated-total annual costs (i.e.,  annualized costs) are
$0.433 billion for strategy I and $2.17 billion for strategy II.  It can be
observed in Table 7-2 that the emission control costs for tank storage are
                                    7-6

-------
               TABLE 7-1.  ESTIMATED TOTAL CAPITAL  INVESTMENT AND  TOTAL  ANNUAL  COST PER
                      UNIT OF WASTE THROUGHPUT BY SOURCE CATEGORY  FOR  TWO  EXAMPLE
                                          CONTROL STRATEGIES3
                                          Total  capital  investment,
                                               S/Mg  throughput0
Total annual cost,
 S/Mq throughput0
Example control strategy"
Source category
Drum storage
Dumpster storage
Storage tank
Quiescent surface impoundments
Quiescent treatment tanks
Aerated/agitated tanks and impoundments
Wastepi les
Landf i 1 Is
Waste fixation
Inc i nerat i on"
Land treatment
Other treatment®
Spi 1 Is
Load i ng
Equipment leaks'
I
0
6.3-26
9.7-28
1.9-2.6
0.22-1.2
0.41-2.9
09
09
0
0
09
0
0
0.49-0.89
0.02-48.0
II
49-91
6.3-26
9.7-28
0
0
0
09
09
0
290-480
09
3.6-66
0
0.49-0.89'
0.02-48.0
Example control strategy''
I
0
2.1-9.9
4.9-15.0
0.87-4.8
0.14-0.68
0.26-1.7
09
09
6.1
0
09
0
0
0.09-0.17'
<0. 01-1. 60
II
21-38
2.1-9.9
4.9-16.0
0
0
0
09
09
0
110-150
09
(23.0)^-21
0
0.09-0.17'
<0. 01-10.0
8Total capital investment includes all  costs to purchase equipment,  direct installation charges,  and
 indirect installation charges.  Total  annual  cost is the sum of the annual  operating cost and the
 annualized capital costs.  All costs are in January 1986 dollars.
''The two example control strategies are described in Chapter 5.0.   The first strategy depends
 predominantly on the use of add-on controls,  and the second strategy offers an  option of  add-on
 controls or organic removal.
cThe unit costs were obtained from information presented in Appendixes H and I.   Where a cost range is
 given, the range represents cost variations due to differences in  waste composition.  Model  waste
 compositions for which costs were derived are presented in Appendix C.
^Incineration represents alternatives to organic removal from the waste organic  sludges and organic-
 containing solids.
eOther treatment represents organic removal  processes such as steam distillation.   Vehts from these
 processes would be regulated under the treatment, storage, and disposal facility  (TSDF) air  standards
 for fugitive emissions and process vents proposed February 7, 1987.
'This source category wiI I be regulated by the proposed TSDF air standards for equipment leaks and
 process vent controls.
9As noted in Chapter 5.0, these source categories are regulated by  the Resource  Conservation  and  Recovery
 Act land disposal  restrictions.
nParentheticaI numbers denote negative values, which means there is a net credit for applying the control
 to the source category.
'Example control  strategies address only controlling drum  loading.

-------
                     TABLE 7-2.  ESTIMATED NATIONWIDE TOTAL CAPITAL INVESTMENT AND TOTAL ANNUAL COST
                                           FOR TWO EXAMPLE CONTROL STRATEGIES3
i
oo
Total capital investment, $10®
Example control strategy'5
Source category
Drum storage
Dumpster storage
Storage tank
Quiescent surface impoundmentsc
Quiescent treatment tanksc
Aerated/agi tated tanks and impoundments0
Wastepi les
Landf i 1 Is
Waste f i xati onc
Inci nerati on^
Land treatment
Other treatment6
Spills
Load i ng
Equipment leaks*
Total
I
0
13.2
162
839
6.89
46.9
09
09
0
0
09
0
0
16.01h
0
1,084
II
1,640
13.2
235
0
0
0
09
09
0
370
09
4,920
0
16.01h
0
7,190
Total annual
1 cost, $106
Example control strategy'3
I
0
5.103
49.5
349
1.44
24.6
09
09
0
0
09
0
0
3h
0
433
II
678
5.103
89.6
0
0
0
09
09
0
143
09
1,250
0
3"
0
2,170

        aTotaI  capital investment includes all costs to purchase equipment, direct installation changes, and
         indirect installation charges.  Total annual cost is the sum of the annual operating cost and the
         annuaIized capital costs.  All costs are in January 1986 dollars.
        ^Example control  strategy I is based predominantly on the use of add-on emission controls.  Example
         control  strategy II also requires the use of add-on emission controls unless the waste stream is
         pretreated by organic removal.  The example control strategies are described in Chapter 5.0.
        cln example control strategy II, the costs for these source categories are included in the organic removal
         costs.
        ^Incineration represents alternatives to organic removal from the waste organic sludges and organic-
         containing solids.
        eOther treatment represents organic removal  processes such as steam distillation.  Vents from these
         processes would be regulated under the treatment, storage, and disposal (TSDF) air standards for
         fugitive emissions and process vents proposed February 7, 1987.
        f"This source category will be regulated by the proposed TSDF air standards for equipment  leaks and
         process vent controls.
        9As noted in Chapter 5.0, these source categories are regulated by the Resource Conservation and Recovery
         Act  land disposal  restrictions.
        "Example control  strategies address only controlling drum  loading.

-------
 higher  for  control  strategy  II  than  for control  strategy  I,  even  though the
 specified controls  are  the same for  both strategies.   The difference  is
 attributable  to  the fact  that,  when  organic  removal  process  are  used  in a
 facility that has no drum or tank storage present  before  control,  it  is
 assumed that  tank storage capacity must be added to  the facility  to serve
 the  organic removal  process  equipment.
      The nationwide cost  information presented  in  this chapter provides two
 means of comparing  control  strategy  capital  and  annual costs.  Other  means
 of comparing  strategies are  discussed in Chapter 5.0.  Section 5.5
 describes a methodology for  ranking  control  strategies according  to the
 relative health  and environmental  benefits achieved  by the strategies.
 7.3   COST EFFECTIVENESS OF CONTROL STRATEGIES
      Table  7-3 shows the  cost effectiveness  of  the two example control
 strategies.   The cost effectiveness  of  a control strategy is defined  as the
 total annual  cost of applying controls  to all emission sources covered by
 the  strategy  divided by the  total  emission reduction that would be
 achieved.   Total annual  cost is the  annual operating cost plus the annual
 cost  of capital  required  to  purchase and install the controls.  As shown,
 the  cost effectiveness  of example strategy I and example  strategy  II  is
 $294/Mg and $l,460/Mg of  organic emission reduction, respectively.
      Only a single  aggregate cost  effectiveness  is presented for  each TSDF
 control strategy.   The  cost  effectiveness of controlling  specific  emission
 source categories covered by a  strategy (e.g.,  the cost effectiveness of
 controlling storage tanks)  is not  presented.  This is because emissions
 from  TSDF sources are interrelated in many strategies and, consequently, it
 is potentially misleading to estimate the cost  effectiveness on an emission
 source category  basis.
      For example, covering only the  first of several TSDF waste management
 units (emission  sources)  in  series will  reduce  organic emissions  from the
 unit that is  covered  but  may increase the emissions  from  the uncovered
 units downstream, resulting  in  no  change in  total  facility emissions.  The
 cost effectiveness  of controlling  the first  unit may look attractive  in
 isolation,  but,  as  a  practical  concern,  reduction  in total facility
 emissions would  not  be  achieved unless  units downstream were controlled as
well.
                                    7-9

-------
           TABLE 7-3.  NATIONWIDE TSDF COST EFFECTIVENESS OF TWO
                        EXAMPLE CONTROL STRATEGIES3


                                              Cost effectiveness
Example control strategy^                   $/Mg emission reduction


     I                                                294

     II                                              1,460


aThis table presents total  annual  costs of control divided by organic
 emission reductions, i.e.,  cost-effectiveness.

^Example control  strategy I  is based predominantly on the use of add-on
 emission controls.   Example control strategy II also requires the use of
 add-on emission  controls unless the waste stream is pretreated by organic
 removal.  The example control strategies  are described  in Chapter 5.0.
                                   7-10

-------
7.4  REFERENCES
1
5.
6.
7.
U.S. Environmental Protection Agency.  EAB Control Cost Manual (Third
Edition), Section 2:  Manual.Estimating Methodology.  Office of Air
Quality Planning and Standards, Economic Analysis Branch.  February
1987.
Vatavuk, W. M.,  and R. B. Neveril
October 6, 1980.

Vatavuk, W. M.,  and R. B. Neveril
November 3, 1980.

Vatavuk, W. M.,  and R. B. Neveril
December 29, 1980.

Vatavuk, W. M.,  and R. B. Neveril
May 18, 1981.

Vatavuk, W. M.,  and R. B. Neveril
January 24, 1983.

Vatavuk, W. M.,  and R. B. Neveril
April 2, 1984.
Chemical  Engineering.   165-168.


Chemical  Engineering.   157-162.


Chemical  Engineering.   71-73.


Chemical  Engineering.   157-162.


Chemical  Engineering.   131-132.


Chemical  Engineering.   97-99.
8.   Pope-Reid Associates, Inc.  Alternative Waste Management Technology
     Cost Estimates for the California List Land Disposal Restrictions.
     Prepared for U.S. Environmental Protection Agency, Washington, DC.
     May 1987.  p. 22-28.

9.   Mahoney, W. D. (ed.).  Means Construction Cost Data 1986.  Kingston,
     MA, R. S. Means Co., Inc.  1985.

10.  Research Triangle Institute.  Cost of Volatile Organic Removal and
     Model Unit Air Emission Controls for Hazardous Waste Treatment, Stor-
     age, and Disposal Facilities.  Draft.  Prepared for U.S. Environmental
     Protection Agency, Office of Air Quality Planning and Standards.
     Research Triangle Park, NC.  October 24, 1986.

11.  Reference 8.
                                   7-11

-------